JMEPEG (2015) 24:4267–4282 ÓASM International

DOI: 10.1007/s11665-015-1740-6 1059-9495/$19.00

Characterization of Tensile Properties, Limiting Strains,

and Deep Drawing Behavior of AA5754-H22 Sheet
at Elevated Temperature
Sudhy S. Panicker, Har Govind Singh, Sushanta Kumar Panda, and Richard Dashwood

(Submitted July 9, 2015; in revised form August 24, 2015; published online October 26, 2015)

Automotive industries are very much interested in characterization of formability improvement of alu-
minum alloys at elevated temperatures before designing tools, heating systems, and processing sequences
for fabrication of auto-body panels by warm forming technology. In this study, tensile tests of AA5754-H22
aluminum alloy were carried out at five different temperatures and three different strain rates to investigate
the deformation behavior correlating with Cowper-Symonds constitutive equation. Laboratory scale warm
forming facilities were designed and fabricated to perform limiting dome height and deep drawing tests to
evaluate forming limit strains and drawability of sheet metal at different tool temperatures. The forming
limit strain and dome height improved significantly when both the die and punch were heated to 200 °C.
Remarkable improvement in deep drawn cup depth was observed when die and punch temperatures were
maintained at 200 and 30 °C, respectively, producing a non-isothermal temperature gradient of approxi-
mately 93 °C across the blank from flange to center. The forming behavior at different isothermal and non-
isothermal conditions were predicted successfully using a thermo-mechanical FE model incorporating
temperature-dependent properties in Barlat-89 yield criterion coupled with Cowper-Symonds hardening
model, and the thinning/failure location in deformed cups were validated implementing the experimental
limiting strains as damage model.

hydroforming (Ref 3), warm forming (Ref 4), warm hydro-

Keywords AA5754-H22, deep drawing, FE simulation, limiting
dome height, warm forming forming (Ref 5), and gas forming (Ref 6) are being introduced
to enhance their formability. The warm forming technique is a
promising and oldest technique, in which the modification of an
existing tooling is not as complicated when compared to other
methods. The use of AA5754 sheets are sometimes limited to
1. Introduction inner body panels because of poor surface finish due to
undesirable stretch markings on the surface when formed at
room temperature. This is because of Portvein-Le Chatelier
The low carbon steel was prevalent in the manufacturing of
(PLC) effect which is a macroscopic indication of dynamic
automotive body components because of their cost effective-
strain aging, and this is clearly visible as a serration flow in the
ness and good formability at room temperature. The necessi-
uniaxial tensile testing results (Ref 7). It has been demonstrated
tation of improved safety, fuel efficiency, and better properties
by Bolt et al. (Ref 8) that this defect disappears when the
of formed parts compelled automotive manufacturers to
component is deformed at warm conditions. The uniaxial
produce lightweight components without compromising the
tensile test performed by Ozturk et al. (Ref 9) on AA5754
design and strength. Therefore, the material selection scenario
showed an increase in total elongation without serration at
is changing to include lighter materials such as aluminum,
cryogenic temperatures and warm conditions. The strain rate
magnesium, and titanium alloys, and metal matrix and polymer
hardening effect causes its increase in formability at warm
composites. Among different commercially available aluminum
conditions while increased elongation at cryogenic temperature
alloys, the strain hardenable AA5754 is extensively used in
is associated with work-hardening phenomena. The critical role
automotive industry because of its high strength to weight ratio,
of strain rate sensitivity on warm formability of three aluminum
excellent corrosion resistance even in saline atmosphere, and
alloys-AA5182, AA5754, and AA6111-T4 was investigated by
very high toughness even at temperature near to absolute zero
Li and Ghosh (Ref 10). The uniaxial tensile tests were
(Ref 1). The major hindrance in using aluminum alloys to
performed in the temperature range of 200-350 °C with
fabricate critical auto-body components is their inferior forma-
varying strain rates of 0.015-1.5 s1, and it was observed that
bility at room temperature when compared to low carbon steels.
slow strain rate at elevated temperature provided a substantial
Different techniques such as electromagnetic forming (Ref 2),
improvement in post-uniform elongation and true fracture
strain. It was also reported that the warm formability of 5xxx
Sudhy S. Panicker, Har Govind Singh, and Sushanta Kumar series alloys was superior to precipitation hardened 6xxx series
Panda, Department of Mechanical Engineering, Indian Institute of alloys. Jain et al. (Ref 11) reported that reduction in die profile
Technology, Kharagpur, Kharagpur 721302, India; and Richard radius resulted in decreased limiting drawing ratio (LDR) and
Dashwood, Warwick Manufacturing Group, University of Warwick, increased the tendency of ironing and subsequent fracture. The
Coventry CV4 7AL, UK. Contact e-mail: sushanta.panda@mech.
effect of blank holding force (BHF) on warm deep drawing of
iitkgp.ernet.in.

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4267

magnesium alloy at elevated temperature was studied by coupled thermo-mechanical FE simulation of both stretch
Yoshihara et al. (Ref 12). The localized cooling at blank center forming and deep drawing processes were also performed. The
region was developed, and a drastic improvement in LDR predicted LDHs, strain distribution, deep drawing cup depth,
emphasizing the need of proper temperature distribution across thinning development, and earing profile were validated with
the blank was proposed. The deep drawing experiments on the experimental data.
AA5754-O aluminum alloy sheet by Palumbo and Tricarico
(Ref 13) showed that temperature gradient on the blank has a
significant effect on LDR for warm deep drawing. The
experimental results were compared by performing FE analysis
2. Materials and Methods
in ABAQUS with a fully thermo-coupled model. The influence
of punch speed on the formability was also emphasized. FE The aluminum alloy—AA5754 in H22 temper condition
validation of deep drawn rectangular cup for AA5182 sheet at was selected in the present study. The 1.0-mm-thick sheet metal
warm conditions was performed by Kim et al. (Ref 14). The was obtained from the manufacturer after rolling and subse-
effect of temperature gradient on the failure location was quent partial annealing, and the chemical composition is
studied by measuring the strain distribution and thickness presented in Table 1. The presence of Mg and Mn solute
distribution. Sheet metal forming essentially undergoes biaxial atoms in this alloy increases strain hardening due to the stress
deformation rather than tensile elongation (Ref 15). Therefore, field developed in the crystal around these atoms, which
Li and Ghosh (Ref 16) studied the biaxial deformation behavior hinders dislocation propagation and/or multiplication. Recently,
of AA5182, AA5754, and AA6111-T4 alloys and reported that the automotive industries are paying attention to enhance the
forming limit diagram (FLD) evaluated at 250 °C for automo- application of this non-heat treatable Al-Mg alloy in manufac-
tive aluminum alloy sheet were comparable to low carbon turing of light-weight body panels without compromising the
aluminum killed steel at room temperature. They achieved crashworthiness.
higher part depth without much loss in post-forming yield
strength by keeping the punch temperature 50 °C lesser than 2.1 Tensile Testing
that of die. Naka et al. (Ref 17) performed stretch forming The tensile properties such as yield strength (YS), ultimate
operation for fine-grained AA5083 alloy sheet at temperature tensile strength (UTS), and % elongation of AA5754-H22 were
range of 20-300 °C with varying deformation rate of 0.2- evaluated using sub-size specimens as per ASTM E8 standards
200 mm/min. A significant rise in limiting strain level was (Ref 23). The tests were conducted in 0°, 45°, and 90° with
noticed at 300 °C condition with lowest deformation rate of 0.2 respect to rolling direction (RD) at five different temperatures
mm/min. 30, 100, 175, 200, and 250 °C, and three different strain rates
The formability prediction and faster optimisation of sheet 0.001, 0.1, and 0.16 s1. The true stress-true strain responses in
metal forming process is possible through finite element the plastic flow regime were correlated with Cowper-Symonds
modeling with proper selection of process parameters. Selec- constitutive equation (Eq 1) to describe the influence of strain
tion of an appropriate hardening rule and yield criterion are hardening and strain rate sensitivity.
crucial in FE simulation of aluminum alloy sheet as discussed 2 !1 3
elsewhere (Ref 18-22). In 1989, Barlat and Lian (Ref 18)
s p 4
e_ peff p
proposed a yield criterion that describes the behavior of
p p
ry ðeeff ; e_ eff Þ ¼ ry ðeeff Þ 1 þ 5; ðEq 1Þ
C
orthotropic sheet metals exhibiting planar anisotropy and
subjected to plane stress condition. The influence of yield
criteria on formability prediction of Al-5% Mg alloy was where ry is the yield stress depending on the effective plastic
described by Yoon et al. (Ref 19). The true stress-true strain strain epeff and effective plastic strain rate e_ peff . The terms, C
data obtained from bulge test were curve fitted to Voce and p are strain rate sensitivity parameters. The quasi-static
hardening rule for prediction of formability. The finite element yield stress, rsy , is related with plastic strain by following
modeling of stamp hydroforming process simulations using power law relation (Eq 2):
Barlat-89 model predicted accurate results when compared with
rsy ðepeff Þ ¼ Kðepeff þ eyeff Þn ; ðEq 2Þ
experimental results done by Zampaloni et al. (Ref 20). The
variation of yield locus with temperature for AA5083-O where eyeff is the initial yield strain and it can be described by
aluminum alloy was experimentally determined by Naka Eq 3.
et al. (Ref 21).
 ð1=ðn1ÞÞ
There were no open literature available on analyzing the E
influence of temperature in the formability of AA5754-H22 eyeff ¼ : ðEq 3Þ
K
aluminum alloy imposing non-isothermal conditions. AA5754
aluminum alloy in H22 temper condition was chosen because In Eq 3, E signifies the YoungÕs modulus. The strain rate
of its relative high strength and ductility when compared to sensitivity parameters C and p and the hardening parameters K
other tempered form. Therefore, in this research, an attempt has and n were determined at different temperatures after fitting the
been made to get an insight into the warm temperature Eq 1 in the uniaxial true stress-true strain response (between the
formability of the material by conducting a series of laboratory
scale tensile testing, stretch forming, and deep drawing Table 1 Chemical composition of as received AA5754-
experiments. Stretch forming experiments comprised the eval- H22 alloy
uation of limiting dome height (LDH) at room temperature and
isothermal elevated temperature at 200 °C. Deep drawing tests Element Mg Mn Fe Si Al
were conducted at both isothermal and non-isothermal condi-
wt. (%) 2.60-3.20 0.50 0.40 0.40 Balance
tions by heating the die and punch independently. Fully

4268—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

Fig. 1 Double-action hydraulic press with heating arrangements to conduct warm forming experiments

YS and UTS). The Lankford anisotropy coefficient or plastic embedded in die, binder, and punch, respectively. Thus, a pre-
strain ratio (R value) of the material was evaluated at different set temperature of the die, binder, and punch were achieved
temperatures to assess the variation with temperature. The with the aid of cut-off switches in the control panel. Insulation
specimens were elongated to 80% of longitudinal strain comprising syndonia sheet and glass wool were placed between
corresponding to UTS for determining the R values (Ref 24). binder and pressure pad to arrest heat transfer. To obstruct heat
The final width and length of the deformed specimens were transfer to the load cell, the coupler connecting the punch and
measured and R values were computed by the following the primary ram was modified with a water jacket. All the
equation: heaters were switched off during room temperature experi-
 ments. The process sequences adopted in the present warm
ew ew lnðwf w0 Þ forming experiments consisted of five basic steps (a-e) enlisted
R¼ ¼ ¼  ; ðEq 4Þ
et ðew þ el Þ lnðl0 w0 lf wf Þ below:
where ew , et ; and el are true strain in width, thickness, and (a) The binder was moved down and touched with the die,
length respectively, w0 and wf are initial and final width, l0 and the exposed top and side surfaces were covered with
and lf are initial and final gauge length. The average normal the glass wool to diminish convective heat transfer.
anisotropy of the material was determined using expression (b) The heaters were switched on to achieve required tool-
 ¼ 1 ðR0 þ 2R45 þ R90 Þ, where R0 ; R45 ; and R90 are the plas-
R 4 ing temperature and the binder was moved up to place
tic strain ratio along rolling, 45°, and transverse direction of and align the blank on the die.
the sheet metal. The 30 °C temperature was referred as room (c) Adequate blank holding pressure was applied to sup-
temperature, and the same is followed in the subsequent sec- press the wrinkling in the component.
tions. (d) The punch was moved toward the blank and appropriate
dwell time was provided to establish steady temperature
2.2 Formability Tests in the blank center.
2.2.1 Warm Forming Test Rig. The stretch forming and (e) At last, the blank was deformed by moving the punch at
deep drawing experiments were performed using a double- a speed of 20 mm/min and the test was stopped as soon
acting hydraulic press of 50 ton capacity of both primary and as necking/failure observed through a mirror placed be-
secondary ram. Fig. 1 demonstrates the essential parts of the low the die.
press where the formability test rig was mounted. Punch was
connected to the primary ram, and binder was attached to the All the experiments were planned in dry condition without
pressure pad on secondary ram. The die was fastened on the application of lubrication. Henceforward, the die temperature
fixture placed on the press bed. A load cell with linear variable means temperature of both the die and binder.
displacement transducer (LVDT) situated at the top of primary 2.2.2 Limiting Dome Height Test. The tooling assembly
ram helped in capturing the load-displacement data throughout including the punch and dies with heaters for the LDH testing is
the deformation process. In-situ arrangements were provided to shown in Fig. 2(a). The circular drawbead was designed at a
heat die, binder, and punch independently through resistance radius of 36 mm on both the dies to clamp the blank by
heaters, each of 350 W capacity. The temperatures of tooling application of blank holding force. Hence, only the portion of
were displayed on control panel with the help of thermocouples the sheet metal exposed to the die cavity was stretched by the

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4269

Fig. 2 Schematic of different laboratory scale warm forming test setup developed: (a) stretch forming setup, and deep drawing setup with (b)
punch having heater and (c) punch having water cooling provision

Fig. 3 Different geometry specimen with dimensions for experimental determination of limiting strains to construct forming limit diagram

50-mm-diameter hemispherical punch without drawing of monitored by placing a thermocouple. Several experimental
flange materials. The maximum punch displacement at the trials were carried out to achieve proper blank holding force
onset of necking/failure in the blank is identified as the LDH. and blank diameter so that deeper cups could be drawn without
Several blanks having dimensions as shown in Fig. 3 were used wrinkling. After a number of trial experiments, BHF and initial
in the test to deform the material in different strain paths. In blank diameters were selected as 1.5 kN and 110 mm,
order to obtain the biaxial strain path, circular blank of 100-mm respectively. As the material strength alters at elevated tem-
diameter was fabricated and the bulge test was carried out. The peratures, the applied BHF was scaled accordingly. Further, to
circle grid analysis (CGA) was performed to measure the major investigate the influence of the die and punch temperature on
and minor strain at different location on the outer surface of the the cup height and thickness distribution, the following four
deformed blank. The stretch forming experiments were con- deep drawing experiments were performed: 1st—both the die
ducted both at room temperature and elevated temperature and punch at room temperature, 2nd—both die and punch at
(200 °C), but the bulge test at elevated condition was not 200 °C, 3rd—punch at 120 °C and die at 200 °C, and
performed due to the setup limitation. For isothermal (200 °C) 4th—punch at 30 °C and die at 200 °C.
LDH test, the blank placed between the dies was held for 10
minutes in order to uniformly heat the blank to the desired
2.3 Finite Element Modeling
temperature.
2.2.3 Deep Drawing Test. The schematic for isothermal The thermo-mechanical simulation of the stretch forming
deep drawing setup with provisions of inserting electric heater and deep drawing process was executed using the LS-971
is shown in Fig. 2(b). Another punch with an arrangement of version of the explicit dynamic finite element code LS-
constant water circulation in the punch to conduct non- DYNA3D (Ref 25). The mechanical and thermal parts of the
isothermal deep drawing operation is as shown in Fig. 2(c). problem are solved independently. The mechanical part of
The water-circulated cooler punch helped in maintaining a problem is solved using explicit dynamic solution scheme,
temperature gradient across the blank during the deformation. whereas the thermal part is solved using implicit scheme. The
The variation in the blank center temperature (BCT) was bottleneck in computation time is the explicit mechanical time

4270—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

step which should be less than the time required for a stress As the thickness of sheet material is negligible compared to
wave to cross the smallest element [Courant criterion (Ref 26)]. other dimensions, deformation considered to be in plane stress
FE models of stretch forming and deep drawing processes were state and hence Barlat-89 yield criterion as proposed by Barlat
developed with the pre-processor Dynaform-5.9.2. Taking and Lian (Ref 18) can be described as
advantage of symmetry, only quarter model was made in order
to reduce computational time. The symmetry boundary condi- rM ¼ ajK1 þ K2 jM þajK1  K2 jM þcj2K2 jM ;
2 ðEq 5Þ
tions were imposed along the sides of the blank in the XZ and where r  is yield stress and M is the Barlat yield surface
YZ planes. All tooling parts treated as rigid bodies assigning exponent, and as aluminum has FCC structure, M ¼ 8 was
material model *MAT_020 with four-node rigid shell element selected. The stress invariants, K1 and K2 found out as shown
meshing. This will help in increasing computational efficiency in Eq 6.
as there will be no requirement of storage allocation for
deformation history (Ref 25). The four-node quadrilateral "  #1=2
rx þ lry rx  lry 2 2 2
Belytschko-Tsay shell elements were used to mesh the K1 ¼ ; K2 ¼ þq rxy ; ðEq 6Þ
deformable blank of thickness 1 mm (Ref 27) with five 2 2
through thickness integration points. To suppress hourglassing,
where rx ; ry ; and rxy are plane stress components in ortho-
i.e., zero stress state during deformation, the default LS-
tropic axes. The anisotropy material parameters a, c, l, and q
DYNA3D hourglass control technique was implemented. The
can be related to LankfordÕs parameters R0 ; R45 , and R90 as
shell element formulation was done using adaptive remeshing
described by Barlat and Lian (Ref 18).
scheme with level-4 refinement for better accuracy in results
Now, taking into account only normal anisotropy, i.e.,
with less number of elements in the model. In order to further  the Eq 5 reduces to an
reduce computational expense, time-scaling method was assuming R0 = R45 = R90 = R,
adopted and rate dependent properties were scaled accordingly. equation in terms of principal stresses.
The punch velocity during simulation was set 1000 mm/s, and 2 2 
2R
the BHF was calibrated since quarter model had been used. For rM ¼
2  jr1 jM þ  jr2 jM þ M
 jr1  r2 j :
1þR 1þR 1þR
LDH test model, an analytical drawbead was assigned in the
ðEq 7Þ
binder at 36 mm radius. The input values of static coefficient of
friction at tool blank interfaces for different operating condi- If the yield function exponent M = 2, then Eq 7 transform to
tions are listed in Table 2. It has been reported that the Hill-48 (Ref 29) anisotropic yield criterion considering only
coefficient of friction between blank and punch reduces with normal anisotropy.
increase in temperature (Ref 28). The surface to surface thermal
2 2 2R
penalty stiffness-based contact algorithm was used to model the r2 ¼ 2 2 2
contact between the blank and the tooling. A constant thermal
2
1þR r1 þ 1 þ R
 r2 þ 1 þ R ðr1  r2 Þ : ðEq 8Þ

gap conductance of 900 W/mK given for the tool and blank Deep drawing simulation using Hill-48 yield criterion were
interface (Ref 13). Thermal properties used for blank material also done for comparison purpose.
(aluminum) and tools (steel) are summarized in Table 3. The To incorporate strain rate effect in the simulation, Cowper-
punch was initially brought into contact with the blank, and Symonds constitutive equation (Eq 1) (Ref 30) was employed
subsequently it was kept touching at this position for five in the present FE model. It was found that the Lankford
seconds to allow heat transfer. As soon as near to steady state  hardening param-
anisotropy coefficients ( R0 ; R45 , R90 ;and R),
condition was reached, the punch was ramped up to its set eters (K and n-value), and strain rate sensitive parameters C and
velocity to deform the material. This technique was used to p were depending on the temperature. The details are discussed
match the experimental conditions. The reason for allowing the
heat transfer to reach a steady state or near steady state
condition was to reduce the variability in the experiments, Table 3 Thermal properties of blank and tooling used in
giving more repeatable results and also making them easier to FE model (Ref 31)
simulate. In the post-processing stage, the forming limit
diagram developed at room temperature and 200 °C were used Blank Rigid tools
to predict the initiation of necking/crack. Properties (Aluminum) (steel)
The deformable blanks were modeled using Barlat-89 yield
function which take into account both planar and normal Heat Capacity (J/kgK) 904 450
anisotropy of aluminum alloy sheet material, defined and Thermal Conductivity (W/mK) 220 70
Density (kg/m3) 2710 7850
executed in LS-DYNA as material type *MAT_036 (Ref 25).

Table 2 Friction coefficient used in FE simulation

Process Operating temperature Coefficient of friction

Stretch forming Room temperature 0.33

200 °C 0.23
Deep drawing Room temperature 0.33
Non-isothermal (Die=200 °C, Punch=30 °C) 0.23 (at die-blank surface interface);
0.33 (at punch-blank surface interface)
200 °C 0.23

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4271

in section 3.1, and these temperature-dependent parameters are 250 °C had been compared as shown in the Fig. 4(a). The
correlated with temperature to be coupled with Cowper- serrated flow was observed in the material during slow strain
Symonds equation and Barlat-89 model. rate (0.001 s1) testing at lower temperatures (both 30 and
100 °C). It may be due to temporary hindrance of dislocations
at high energy state grain boundaries with the presence of
solute Mg-atoms. The presence of solute atoms around the
3. Results and Discussion grain boundaries may have created stronger barrier, and load
bearing capacity increases temporarily with a sharp drop in load
3.1 Tensile Properties and Cowper-Symonds Constitutive once these barriers are torn away (Ref 31). However, the
Equation serrated flow was observed to diminish while uniaxial defor-
Engineering stress-strain response of AA5754-H22 alloy mation at higher strain rate and elevated temperature, as shown
specimens obtained from uniaxial tensile test along the rolling in Fig. 4(a) and (b). The velocity of mobile dislocations during
direction at a strain rate of 0.001 s1 under five different plastic deformation increases at higher strain rate, and this helps
temperatures: room temperature (30 °C), 100, 175, 200, and in reducing the waiting time of the dislocations at grain

Fig. 4 (a) Effect of temperature on engineering stress-strain response at constant strain rate of 0.001 s1 and (b) Strain rate effect on tensile
deformation with validation of Cowper-Symonds constitutive equation

Table 4 Mechanical properties of AA5754-H22 alloys at five different temperatures for a quasi-static strain rate of
0.001 s21
Temperature, °C YS, MPa UTS, MPa Uniform elongation, % Post-uniform elongation, % Total elongation, %

30 87.00 185.00 20.50 4.00 24.5

100 100.00 201.00 24.00 3.70 27.7
175 86.00 155.00 18.50 43.50 62.0
200 82.00 127.00 14.10 49.90 64.0
250 72.50 102.00 11.30 75.20 86.5

Fig. 5 Variation of Cowper-Symonds parameters with temperature:(a) K and n-value, and (b) C and p value

4272—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

boundaries. Hence, a weak barrier of solute atoms around the variation in Cowper-Symonds material parameters at different
grain boundaries is easily torn away by the movable dislocation temperature was correlated as shown in Fig. 5(a) and (b). It can
(Ref 30). It was noticed that both the YS and UTS increased at be seen that the n-value decreased with the increase in
100 °C, and similar results were observed by previous deformation temperature, and this is attributed to the reduction
researchers (Ref 7). It was found to be related with the in uniform elongation. Similarly, the C and p values were
predominance of dynamic strain aging (DSA) under PLC range obtained at five different temperatures after non-linear regres-
where temperature has negligible effect on flow stress. The YS sion of Cowper-Symonds constitutive equation (Eq 1) using
and UTS were found to decrease with further increase in true stress and true plastic strain data. The variation of C and p
temperature, and a sharp reduction was observed at tempera- values with temperature is depicted in Fig. 5(b). The higher
tures above 175 °C. value of C at room temperature signifies insensitivity of the
Table 4 summarizes the mechanical properties such as YS, material at room temperature which can also be seen in
UTS, and total elongations of the sheet metal at quasi-static Fig. 4(b). The experimental true stress-strain response at room
strain rate of 0.001 s1. It can be inferred from Table 4 that the temperature for 0.001 and 0.1 s1 are almost coinciding, while
reduction in UTS was more sensitive to temperature than YS. a distinguishable reduction in flow stress is observed for lower
The total elongation was observed to increase by 253% with strain rate at 200 °C. The significant rise in post-uniform
increase in deformation temperature from room temperature to elongation may be due to the strain rate sensitivity of the
250 °C, and this reflected the significant improvement in material at elevated temperature which delayed the onset of
formability of this material. Further, the total elongation was strain localization and subsequent failure. Also, both K and n
decoupled into uniform and post-uniform elongation as values reduced at elevated temperature (Fig. 5a), and this
depicted in Fig. 4(a). A gradual decline in uniform elongation helped in quantitative decrease in the flow stress resulting
from 20.5% to 11.3% was observed when temperature was reduction in deformation load. The influence of temperature on
changed from room temperature to 250 °C. However, signif- Lankford anisotropy parameters R0 , R45 ; and R90 is depicted in
icant increase in post-uniform elongation was noticed at Fig. 6, where a polynomial curve fit is used to represent the
elevated temperature beyond 100 °C. The Cowper-Symonds variations. The increase in these Lankford parameters resulted
constitutive equation was used to capture the effect of strain in 144.1% increase in average normal anisotropy parameter ( R) 
rate on the uniaxial deformation behavior of the material. The at 250 °C compared to room temperature. This improvement
signifies the higher resistance to thinning during sheet forming,
which may help in increase in formability of the material at
elevated temperature. The R2 values specified in Fig. 5 and 6 is
the coefficient of determination which signifies how well data
points fit a curve. The stress-strain responses were plotted at
two different temperatures (30 and 200 °C) and strain rates
(0.001 and 0.1 s1) using the modified Cowper-Symonds
constitutive equation (Eq 9) incorporating the temperature
dependent parameters, and the results were validated with
experimental stress-strain response as shown in Fig. 4(b). The
equations for the temperature-dependent parameters
K ðT Þ; nðT Þ; C ðT Þ; pðT Þ; R0 ðT Þ; R45 ðT Þ; R90 ðT Þ; and RðT Þ were
obtained after curve fitting the values of respective parameters
at different temperatures and are enlisted in Table 5.
2 !1 3
 nðT Þ e_ peff pðTÞ
p p p
ry ðeeff ; e_ eff ; T Þ ¼ KðT Þ eeff 41 þ 5: ðEq 9Þ
CðT Þ

3.2 Forming Limit Diagram

The stretch forming experiments were conducted to evaluate
Fig. 6 Variation of Lankford anisotropy parameters with tempera- the LDH of varying specimen geometry as discussed in
ture section 2.2.2 at room temperature and elevated temperature

Table 5 Curve-fit equations of temperature-dependent Cowper-Symonds and Lankford anisotropy parameters

Parameters Curve-fit equation, T in °C Coefficient of determination, R2

K(T) 7E05T3  0.0346T2 + 3.9471T + 259.75 0.982

n(T) 0.2791  6.869E4T 0.991
C(T) 1E + 06e0.55T 0.959
p(T) 2E8T3 + 6E-06T2  0.0014T + 3.1448 0.980
R0(T) 2E8T3 + 4E-05T2  0.0051T + 0.8909 0.993
R45(T) 2E07T3 + 9E05T2  0.007T + 0.5417 0.998
R90(T) 3E08T3 + 4E05T2  0.006T + 0.7981 0.968
 Þ
RðT 1E07T3 + 6E05T2  0.0063T+0.6931 1.000

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4273

(200 °C). Fig. 7 shows CGA to evaluate both the major and estimating the true stretching strains for both major and minor
minor limiting strains from the deformed ellipse lying within a strains ( e1 and e2 ).
critical region of a 100 9 100 mm specimen deformed at room  
tf
temperature. It is important to note that the specimens are eb ¼ ln 1 þ ; ðEq 10Þ
deformed into the convex shape using the hemispherical punch, 2q
and it may have experienced both stretching and bending where tf is the final thickness and it can be expressed as a
during the deformation. Previously, researchers (Ref 32) function of initial thickness t0 and measured true major and
observed that the experimentally determined limit strains were true minor surface strains as shown in Eq 11.
influenced by the bending strains due to the punch curvature.
The dependence of tensile bending strain ( eb ) induced on the tf ¼ t0 expðem1  em2 Þ ðEq 11Þ
convex surface due to the punch curvature ( qÞ can be
expressed by bending equation as shown in Eq 10. Hence, the e1;2 ¼ em1;2  eb : ðEq 12Þ
measured surface strains increased with the increase in bending
strains leading to a higher forming limit diagram. In order to
The true major and minor strains at safe, necking, and
evaluate the forming limit strains correctly, Eq 12 was used to
fracture regions in deformed specimens of different geometries
subtract the bending strain from the measured surface strain (
were measured. All these data were plotted in strain space with
em1;2 ). Thus, the effect of punch curvature was corrected for
minor and major strain as abscissa and ordinate, respectively, as
shown in Fig. 8. The forming limit diagram (FLD) was
constructed to separate the safe zone from the failure zone
consisting of necking and fracture strain data. Further, the
bending strain correction was incorporated in the entire curve to
estimate the corrected FLD. It was observed that the deformed
specimens of 100 9 60 mm and 100 9 40 mm dimensions
provided data closer to plane strain condition. Even though
exact data points were not available for pure plane strain
condition, it was considered that the strain value where the FLD
intersected the major strain axis as FLD0. Mostly, the failure
occurred in plane strain condition while stamping an industrial
complex components due to lack of lubrication and punch
geometry, and further the limiting strains are lowest along plane
strain deformation mode. Hence, several researchers considered
FLD0 as a measure of formability, and sheet forming industries
are interested in improvement in FLD0 at elevated temperature
apart from the complete FLD as a diagnostic tool. The FLDs
evaluated both at room temperature and 200 °C are compared
in Fig. 8. All the stretch formed specimens are shown in
Fig. 9(a) and (b), and the deformed ellipses on these specimens
Fig. 7 Circular grid analysis in a stretch formed 100 9 100 mm were measured to generate the limiting strain data for
specimen constructing the FLDs as discussed above.

Fig. 8 Forming limit diagrams with bending correction: (a) room temperature and (b) elevated temperature at 200 °C

4274—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

Fig. 9 Deformed LDH test specimens for evaluation of forming limit diagrams: (a) room temperature and (b) elevated temperature at 200 °C

Table 6 Limiting dome height comparison of experimental and FE predicted results at room temperature and 200 °C
Dome height at 30 °C, in mm Dome height at 200 °C, in mm

Simulation Results % error

Specimen geometry
(mm 3 mm) Expt. Simulation Results % error Expt. without C&p with C&p without C&p with C&p

100 9 100 14.78 ± 0.08 14.72 0.41 19.06 ± 0.06 14.08 19.01 26.12 0.26
100 9 80 13.86 ± 0.09 13.88 0.14 18.55 ± 0.08 13.89 18.70 25.12 0.81
100 9 60 13.44 ± 0.08 13.00 3.27 16.70 ± 0.07 11.45 17.09 31.44 2.33
100 9 40 16.94 ± 0.07 16.90 0.23 18.48 ± 0.06 13.44 18.75 27.27 1.46
100 9 20 (Hasek) 13.79 ± 0.08 13.80 0.08 17.49 ± 0.09 12.27 17.29 29.89 1.14
(Bulge test) 100 mm diameter 29.61 ± 0.10       

3.2.1 Effect of Bending and Temperature on Forming 3.3 Limiting Dome Height Validation
Limit Diagram. The corrected FLD was shown as dotted line
Table 6 summarizes the comparison of experimental LDH
after neglecting the bending strain in Fig. 8 at both room and
with predicted values for all specimen geometries used in the
elevated temperature of 200 °C. It can be observed that the
experiment. The dome heights were predicted by FE simulation
corrected FLD shifted to a marginally lower position when
implementing the Cowper-Symonds constitutive equations with
compared to the experimentally measured FLD. Simultane-
and without incorporating the strain rate sensitive parameters
ously, it shifted toward the tension-compression region by
(C and p values). It was found that closer dome height as that of
approximately 3.41 and 3.1% at room temperature and 200 °C,
experimental result was predicted at 200 °C when incorporat-
respectively. This shift was due to the induced bending strain
ing the strain rate sensitive parameters in FE modeling,
from the 50-mm-diameter hemispherical punch during LDH
however, the results were under predicted otherwise. These
testing, which was subtracted from the measured strain. As the
LDH results give insight into the influence of strain rate
true bending strain ðeb Þ is not a constant value, rather a function
sensitive parameters to be crucial in helping the increase in
of tf and q, hence the repositioning of corrected FLD in
formability at elevated temperature. However, the LDH results
downward direction was not uniform. However, the shift was
were predicted close to the experimental results without
lower at elevated temperature due to higher dome height of the
incorporating the C and p parameters at room temperature.
specimens.
The error in prediction lies reasonably within 0.14 to 3.27%, in
Significant improvements in the limiting strains were
an acceptable range. These minor variations may be attributed
observed in the entire FLD at elevated temperature of 200 °C
to the assumption of a constant friction coefficient. Also, the
compared to that at room temperature. The FLD at 200 °C can
several rate-dependent properties like thermal conductivity,
be observed to shift in upward direction significantly as shown
convective, and radiation heat transfer coefficient were assumed
in Fig. 8(b), and the FLD0 changed from 0.18 to 0.44 indicating
in FE simulations, and a small error will significantly influence
145% improvement. This enhancement in formability may be
the simulation results due to adoption of the time-scaling
due to the drastic improvement in post-uniform elongation of
method.
the material delaying the onset of strain localisation at elevated
As the material possesses very encouraging ductility the
temperatures. The equi-biaxial, plane strain, and tension-
fracture in all specimens was not sudden, rather after the
compression strain path of 20-mm-width Hasek (Ref 33)
occurrence of visible localized necking. In all FE simulation,
specimens are shown in FLDs. It can be observed that
the necking location was predicted very close to that observed
insufficient availability of data points in pure biaxial deforma-
in the experiments. The comparative representation of exper-
tion at elevated temperature due to experimental limitations
imental and predicted LDH at elevated temperature for biaxial
enforced to do minute extrapolation to plot complete FLD.

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4275

Fig. 10 Experimental and FE predicted LDH comparison after stretch formed at elevated temperature at 200 °C: (a) biaxial specimen and (b)
Hasek specimen

Fig. 11 Strain distribution profile of biaxial specimen deformed at (a) room temperature and (b) 200 °C; and Hasek specimen deformed at (c)
room temperature and (d) 200 °C

and Hasek specimens with failure location is depicted Fig. 10. Fig. 11. The FE predicted results were also incorporated to
It can be found that the failure location from the pole in the validate the results. It can be observed that both the major and
stretch formed components shifted depending on the geometry minor strains are positive values indicating biaxial tensile
of the specimen, e.g., 11 mm in case of Hasek specimen and 20 deformation mode for 100 9 100 mm specimens during LDH
mm in case of biaxial specimens, both deformed at elevated testing both at room temperature and 200 °C (Fig. 11a, b). The
temperature. This may be due to the strain distribution strains were observed to be very well developed in the
developed in the component depending on the geometry and specimens when deformed at elevated temperature due to the
strain path during deformation. higher LDH. The location of peak strains was observed to shift
3.3.1 Strain Distribution Profile. The major and minor away from the pole with rise in temperature due to the
strains were measured on surface of the biaxial and Hasek redistribution of strain depending on the material flow on
specimens, and these data were plotted with respect to the surface of the heated punch. In case of the Hasek specimens,
distance from the pole to represent the strain distribution the minor strain was negative indicating the tension-compres-
profile. The comparison of strain distribution profile of the sion deformation mode due to the lateral drawing of the
deformed cups at room temperature and 200 °C is shown in material during LDH test, as shown in Fig. 11(c) and (d).

4276—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

Fig. 12 Deep drawn cups obtained from initial blank diameter of 110 mm at different die and punch temperature combinations: (a) 30 and
30 °C; (b) 200 and 200 °C; (c) 200 and 120 °C; and (d) 200 and 30 °C

Similar to biaxial tensile deformation, the peak strain increased

significantly at elevated temperature during tension-compres-
sion deformation mode. However, the location of peak strain
did not change due to the specimen geometry with lesser
material at the center of the Hasek specimen. Hence, the failure
was observed consistently closer to the pole compared to that in
biaxial specimen. The FE-predicted strain distribution trend can
be observed to be in good agreement with measured values.

3.4 Deep Drawing Behavior

Deep drawing experiments were performed using 110-mm
circular blanks under four different tooling temperature condi-
tions as described in section 2.2.3. The deep drawn cups
obtained from these four different isothermal and non-isother-
mal conditions are shown in Fig. 12 with the cup depth
indicating the influence of tool temperature. In all the elevated
temperature deep drawing experiments, the die was kept
consistently at 200 °C and the punch temperature was varied
to three different temperatures, i.e., 200, 120 °C, and room
temperature. The cup depth increased marginally from 13.17 Fig. 13 Thickness distribution in experimental deep drawn cups for
mm at room temperature to 13.71 mm when both the punch and different die and punch temperature combinations with close view of
die were heated to 200 °C. Further, the cup depth was increased thinning development
to 15.16 mm when the punch temperature was reduced to
120 °C, and this indicated the influence of non-isothermal
pressive stress. It was also found that too much temperature
condition. However, the circular blank was drawn completely
difference between the die and punch temperature resulted in
with an average cup depth of 42.57 mm when the punch
early failure at cup top corner because of higher temperature
temperature was further brought down to room temperature by
flange material.
water circulation.
The comparison of thickness distributions measured from
3.5 FE Validation of Deep Drawing Process
the bottom center to the top edge of the cup is shown in Fig. 13.
The close view of the thinning development in all the The cup height predicted from the thermo-mechanical FE
experimentally drawn cup is also shown. In all the drawn simulation was compared with the experimental results for all
cups, the maximum thinning took place at the cup bottom the four different isothermal and non-isothermal deep drawing
corner. However, it can be observed that the thinning devel- cases. It can be seen in Fig. 14 that predicted cup heights are in
opment reduces in non-isothermal deep drawing process. better agreement with experimental results within a reasonable
Hence, a bigger cup could be drawn by providing temperature error of 5%. The thickness distribution with maximum thinning
gradient within the blank in warm forming. Also, significant development was predicted very close to the experimental
thickening of 35% in the upper wall was observed in a results, as shown in Fig. 15. The minor change in thickness was
completely drawn cup due to the circumferential/hoop com- observed on the flat bottom portion of the cup due to the

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4277

negligible deformation because of the frictional force between drawn into the die cavity and the material flow was not uniform
the sheet and punch interface from the start of the deep drawing due to the difference in resistance offered by the material in
process. On further punch movement, the flange materials were different direction because of the anisotropic properties. This
mechanism of material flow led to wavy top edge, often
referred as earing, in a completely drawn cup as observed in
non-isothermal deep drawing process. To determine the earing
profile, cup height was measured at different locations of cup
with an increment of 6o angular position from rolling direction.
The earing profile and % earing (i.e., ððH  hÞ=hÞ  100,
where H was cup height at any location and h was the minimum
height) predicted by FE simulation was validated with exper-
imental earing profile. It can be observed that the FE simulation
was able to capture the earing defect in reasonable agreement
with experimental values (Fig. 16a, b). Total four ears having
peak heights at 0°, 90°, 180°, and 270° with respect to the
rolling direction were observed in the present study due to the
positive planar anisotropy (i.e., DR ¼ 12 ðR0  2R45 þ R90 Þ > 0).
3.5.1 Insight into Formability Improvement in Non-
isothermal Deep Drawing. To get insight into the formabil-
ity improvement, finding the temperature distribution in the cup
during deformation is very important. The flow properties of
the material were already found to be influenced heavily on the
temperature during deformation. Further, to validate the FE
model, the blank center temperature (BCT) was measured
during the progress of deep drawing experiment, as this portion
Fig. 14 FE validation of deep drawn cup heights for different die
within the blank was easily accessible for inserting the
and punch temperature combinations
thermocouple. Fig. 17 shows the comparison of FE predicted

Fig. 15 FE validation of thickness distribution in deep drawn cups: (a) room temperature; (b) isothermal (200 °C); and (c) non-isothermal
(punch at room temperature and die at 200 °C)

4278—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

Fig. 16 (a) FE validation of earing profile and pct. earing and (b) comparative demonstration of earing profile in non-isothermal deep drawn
cups

Fig. 17 Validation of blank center temperature (BCT) with temperature distribution across the blank at different deformation steps

and experimental measured BCT with progression of cup decreased as shown in Fig. 17a-c. The transient temperature
height. As per the adopted process sequence in section 2.2.1, distribution profile developed within the blank during defor-
the blank got heated up uniformly to 200 °C after placing mation process can be observed at different time steps. Hence,
between heated die and binder. The temperature at bottom the different regions within the cup were maintained at different
portion of the blank decrease instantly to 107 °C once it came temperature during the deep drawing process, and this helped in
in contact with the water-circulated cooler punch (Fig. 17a), controlling the material flow into the die cavity. The punch
and this established approximately 93 °C temperature differ- corner and flange region temperature was captured from FE
ence across the blank. This is due to the conductive heat simulation at an intermediate cup depth (Fig. 18a), and it was
transfer from the blank to punch before the start of deformation. found that the flange region and the cup corner were at 180 and
In order to establish a steady BCT, the punch was allowed to be 140 °C, respectively. The flow properties of the AA5754-H22
in touch with the blank for about 5 s. It can be observed in alloy were already observed to be significantly influenced by
Fig. 17 that BCT remain almost constant throughout the the temperature, and the strength was found to decrease with
deformation process and similar result is predicted by FE the increase in temperature. The Barlat-89 yield locus incor-
simulation as well. In the flange region two different conduc- porating the temperature-dependent properties was plotted at
tion heat transfer processes took place, and these are (i) from room temperature, 140 and 200 °C to represent the change in
flange to blank center due to significant temperature difference biaxial stress state of the material with respect to different
between them and (ii) from both die and binder to flange due to temperature condition (Fig. 18b). It can be seen that the yield
contact. As the punch progressed down, the first phenomenon locus is bigger for 140 °C and it shrinks drastically at 200 °C.
dominate over second, therefore the flange temperature Hence, the strength of the material at the cup bottom corner

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4279

Fig. 18 (a) Deformed cup at an intermediate step representing the temperature variation and (b) Influence of temperature on Barlat-89 yield lo-
cus for AA5754-H22

Fig. 19 Influence of yield criteria in FE predictions: (a) cup height and (b) thickness distribution

maintained at 140 °C increased to state Q with simultaneous predicting better results than Hill-48 yield criterion. It can be
decrease in strength of flange material to state C. This helped in observed in Fig. 19(b) that different thickness distributions
easy material flow from the flange with simultaneous increase with thinning development along both the rolling and trans-
in the strength of the cup wall in order to draw a bigger cup verse directions were predicted very well by the Barlat-89 yield
successfully. criterion. However, the Hill-48 model could not able to predict
this difference in thickness distribution in different section as it
3.6 Influence of Planar Anisotropy neglected the planar anisotropy.
Thermo-mechanical FE simulation was also carried out
3.7 Effect of Deformation Mode on Thinning Development
using the Hill-48 yield criterion incorporating only normal
anisotropy parameter. The cup height obtained from this yield The susceptibility of further deformation can be identified
criterion neglecting the planar anisotropy could not able to from the evolution of thinning in a formed component. Hence,
predict the wavy ear profile, and hence it was concluded that the pattern of thinning development was compared due to the
the earing defect was due to the planar anisotropy in the influence of different forming modes used in the present
material during non-isothermal deep drawing. The maximum studies, namely stretch forming and deep drawing process. The
and the minimum cup height obtained from the deep drawing true thickness strain variation along normalized distance from
experiments are compared with FE predicted results using both the cup center (or pole) is shown in Fig. 20. The normalized
the Barlat-89 and Hill-48 yield criterion as shown in Fig. 19(a), distance is used to negate the effect of punch diameter, but the
and it can be observed that the Barlat-89 yield criterion is effect of deformation mode on thickness distribution can be

4280—Volume 24(11) November 2015 Journal of Materials Engineering and Performance

 (b) The limiting strains were evaluated by carrying out the
LDH test of different specimen geometries both at room
temperature and elevated temperature (200 °C). The ef-
fect of punch curvature and temperature on the FLD
was established. There was very negligible influence of
bending strain developed due to the hemispherical
punch used in the present LDH tests. However, the FLD
shifted upward drastically at 200 °C with 145%
improvement in the FLD0.
(c) The thermo-mechanical FE model was developed suc-
cessfully coupling the Barlat-89 yield criterion with the
established Cowper-Symonds constitutive equations
incorporating temperature-dependent anisotropy parame-
ters. The LDH was predicted within 5% error compared
to the experimental results. The FE simulation results
suggested that the improvement in LDH at elevated tem-
perature was due to the strain rate sensitivity of the
AA5754-H22 material, which improved the surface
strain distribution and delayed the onset of necking.
Fig. 20 Effect of deformation mode on thinning development in d) The influence of tooling temperature was investigated by
different warm forming processes
conducting deep drawing experiments at different die and
punch temperature combinations to understand the con-
trol of material flow due to temperature gradient. The
observed. In the hemispherical stretch forming, the material
significant improvements in cup heights were observed
exposed above the die cavity gets stretched through biaxial
when the die was heated to 200 °C and the punch was
tensile stress which led to thinning of material. Hence the
maintained at room temperature. The FE simulation of
thickness strain was negative, and the maximum thinning took
this deep drawing condition predicted that a non-isother-
place corresponding to the location of peak surface major
mal temperature distribution was established in the blank
strain. The maximum thinning location during stretch forming
with a maximum temperature difference of 93 °C from
was found to shift away from the pole with increase in
flange to center.
deformation temperature. In deep drawing process, the defor-
(e) The improvement in formability in non-isothermal deep
mation of the blank took place due to the radial tensile and
drawing process was due to the increase in the strength
circumferential compressive deformation mode, and further the
of the cup wall maintained at lower temperature and
blank bends and unbends across the die corner radius while
simultaneous easier flow of flange material held above
flowing into the die cavity. The complete load during deep
the die at higher temperature. This helped in reduction
drawing process was supported by the cup wall and hence
in thinning development at the cup bottom corner, and
thinning was observed consistently at the punch corner due to
hence the 110-mm-diameter blank was deep drawn suc-
strain localization. Further, the thickness distribution and
cessfully. The Barlat-89 anisotropy material model could
thinning development during deep drawing was different
able to predict the earing profile and thickness distribu-
compared to the stretch forming process as observed in
tions in different direction with respect to rolling direc-
Fig. 20. Positive thickness strain indicating thickening of the
tion due to accurate incorporation of temperature
cup wall was noticed at the top of the deep drawn cup due to
dependent planar anisotropy properties. It was also con-
the circumferential compressive stress.
cluded that the thickness distribution and the thinning
location depended on the deformation mode inherited
from the tool geometry and the forming process.
4. Conclusions

The deep drawing and stretch forming test setup were

developed to carry out laboratory scale isothermal and non- Acknowledgment
isothermal formability testing, and the following are the major
Authors are thankful to Mr. Chandan Mondal, Metal Forming
conclusions made from the experimental and FE studies:
Laboratory, Department of Mechanical Engineering, IIT Kharagpur
for his help while conducting sheet metal forming experiments.
(a) The uniaxial tensile test results showed that the
AA5754-H22 alloy was temperature and strain rate
dependent. The total elongation increases from 24.5% at
room temperature to 86.5% at elevated temperature of References
250 °C, and the PLC effect reduced significantly both at
elevated temperature and higher strain rate. The Cow- 1. S. Toros, F. Ozturk, and I. Kacar, Review of Warm Forming of
Aluminum-Magnesium Alloys, J. Mater. Process. Technol., 2008, 207,
per-Symonds constitutive equation was established suc- p 1-12
cessfully by incorporating the temperature-dependent 2. V. Psyk, D. Risch, B.L. Kinsey, A.E. Tekkaya, and M. Kleiner,
hardening parameters (K and n value) and strain rate Electromagnetic Forming—A Review, J. Mater. Process. Technol.,
sensitive parameters C and p. 2011, 211, p 787-829

Journal of Materials Engineering and Performance Volume 24(11) November 2015—4281

 3. S.H. Zhang, Z.R. Wang, Y. Xu, Z.T. Wang, and L.X. Zhou, Recent 18. F. Barlat and K. Lian, Plastic Behavior and Stretchability of Sheet
Developments in Sheet Hydroforming Technology, J. Mater. Process. Metals. Part I: A Yield Function for Orthotropic Sheets Under Plane
Technol., 2004, 151, p 237-241 Stress Conditions, Int. J. Plast., 1989, 5, p 51-66
4. D.V. Wilson, Aluminium Versus Steel in the Family Car—The 19. J.W. Yoon, F. Barlat, R.E. Dick, K. Chung, and T.J. Kang, Plane Stress
Formability Factor, J. Mech. Work. Technol., 1988, 16, p 257-277 Yield Function for Aluminum Alloy Sheets. Part II: FE Formulation
5. S. Mahabunphachai, M. Koc, and J.E. Carsley, Investigations on and Its Implementation, Int. J. Plast., 2004, 20, p 495-522
Deformation Behavior of AA5754 Sheet Alloy Under Warm Hydro- 20. M. Zampaloni, N. Abedrabbo, and F. Pourboghrat, Experimental and
forming Conditions, J. Manuf. Sci. Eng., 2011, 133, p 051007 Numerical Study of Stamp Hydroforming of Sheet Metals, Int. J.
6. J. Jeswiet, M. Geiger, U. Engel, M. Kleiner, M. Schikorra, J. Duflou, R. Mech. Sci., 2003, 45, p 1815-1848
Neugebauer, P. Bariani, and S. Bruschi, Metal Forming Progress Since 21. T. Naka, Y. Nakayama, T. Uemori, R. Hino, and F. Yoshida, Effects of
2000, CIRP J. Manuf. Sci. Technol., 2008, 1(2008), p 2-17 Temperature on Yield Locus for 5083 Aluminum Alloy Sheet, J. Mater.
7. R.C. Picu, G. Vincze, F. Ozturk, J.J. Gracio, F. Barlat, and A.M. Process. Technol., 2003, 140, p 494-499
Maniatty, Strain Rate Sensitivity of the Commercial Aluminum Alloy 22. D.G. Tari and M.J. Worswick, Elevated Temperature Constitutive
AA5182-O, Mater. Sci. Eng. A, 2005, 390, p 334-343 Behavior and Simulation of Warm Forming of AZ31B, J. Mater.
8. P.J. Bolt, R.J. Werkhoven, A.H. Van Den Boogaard, Warm Deep Process. Technol., 2015, 221, p 40-55
Drawing of Aluminium Sheet, 2003, p 2-9 23. ASTM, E8/E8M Standard Test Methods for Tension Testing of
9. F. Ozturk, S. Toros, and H. Pekel, Evaluation of Tensile Behaviour of Metallic Materials, Annu. B. ASTM Stand., 2010, 4(2010), p 1-27
5754 Aluminium-Magnesium Alloy at Cold and Warm Temperatures, 24. S.K. Panda, D.R. Kumar, H. Kumar, and A.K. Nath, Characterization
Mater. Sci. Technol., 2009, 25, p 919-924 of Tensile Properties of Tailor Welded IF Steel Sheets and Their
10. D. Li and A. Ghosh, Tensile Deformation Behavior of Aluminum Formability in Stretch Forming, J. Mater. Process. Technol., 2007, 183,
Alloys at Warm Forming Temperatures, Mater. Sci. Eng. A, 2003, 352, p 321-332
p 279-286 25. Livermore Software Technology Corporation, LS-DYNA Keyword
11. M. Jain, J. Allin, and M.J. Bull, Deep Drawing Characteristics of UserÕs Manual, Version 971, California (2007)
Automotive Aluminum Alloys, Mater. Sci. Eng. A, 1998, 256, p 69-82 26. R. Courant, Variational Methods for the Solution of Problems of
12. S. Yoshihara, K.I. Manabe, and H. Nishimura, Effect of Blank Holder Equilibrium and Vibrations, Bull. Am. Math. Soc., 1943, 49, p 1-23
Force Control in Deep-Drawing Process of Magnesium Alloy Sheet, J. 27. T. Belytschko, J.I. Lin, and T. Chen-Shyh, Explicit Algorithms for the
Mater. Process. Technol., 2005, 170, p 579-585 Nonlinear Dynamics of Shells, Comput. Methods Appl. Mech. Eng.,
13. G. Palumbo and L. Tricarico, Numerical and Experimental Investiga- 1984, 42, p 225-251
tions on the Warm Deep Drawing Process of Circular Aluminum Alloy 28. H. Karbasian and A.E. Tekkaya, A Review on Hot Stamping, J. Mater.
Specimens, J. Mater. Process. Technol., 2007, 184, p 115-123 Process. Technol., 2010, 210, p 2103-2118
14. H.S. Kim, M. Koç, J. Ni, and A. Ghosh, Finite Element Modeling and 29. R. Hill, A Theory of the Yielding and Plastic Flow of Anisotropic
Analysis of Warm Forming of Aluminum Alloys—Validation Through Metals, Proc. R. Soc. London A Math. Phys. Eng. Sci., 1948, 193, p
Comparisons with Experiments and Determination of a Failure 281-297
Criterion, J. Manuf. Sci. Eng., 2006, 128, p 613 30. J. Winklhofer, G. Trattnig, C. Lind, C. Sommitsch, H. Feuerhuber, F.
15. D.E. Green, K.W. Neale, S.R. MacEwen, A. Makinde, and R. Perrin, Barlat, Y.H. Moon, M.G. Lee, Process Simulation of Aluminium Sheet
Experimental Investigation of the Biaxial Behaviour of an Aluminum Metal Deep Drawing at Elevated Temperatures, 2010, p 927-934
Sheet, Int. J. Plast., 2004, 20, p 1677-1706 31. N. Abedrabbo, F. Pourboghrat, and J. Carsley, Forming of Aluminum
16. D. Li and A.K. Ghosh, Biaxial Warm Forming Behavior of Aluminum Alloys at Elevated Temperatures - Part 2: Numerical Modeling and
Sheet Alloys, J. Mater. Process. Technol., 2004, 145, p 281-293 Experimental Verification, Int. J. Plast., 2006, 22, p 342-373
17. T. Naka, G. Torikai, R. Hino, and F. Yoshida, The Effects of 32. P.L. Charpentier, Influence of Punch Curvature on the Stretching
Temperature and Forming Speed on The Forming Limit Diagram for Limits of Sheet Steel, Metall. Trans. A, 1975, 6, p 1665-1669
Type 5083 Aluminum-Magnesium Alloy Sheet, J. Mater. Process. 33. V. Hasek, Research and Theoretical Description Concerning the
Technol., 2001, 113, p 648-653 Influence on the FLDs, Blech Rohre Profile, 1978, 25, p 213-220

4282—Volume 24(11) November 2015 Journal of Materials Engineering and Performance