Professional Documents
Culture Documents
A R T I C L E I N F O A B S T R A C T
Keywords: A phenomenon that when the sheet metal flows through drawbead, the limit strains of sheet metal lie above the
Sheet metal forming forming limit diagram indicating fracture in theory, while it is safe in practice, violates the normal forming limit
Formability curve has been found. In order to explain the phenomenon, a series of tests were conducted. In the first part, the
Drawbead conventional uniaxial tensile test was employed as a reference. In the second part, specimens were pulled
Pre-strain
through various drawbead inserts several times, and then cut to the uniaxial tensile test specimens and tested by
the conventional uniaxial tensile test. In the third part, specimens were pulled through drawbead inserts many
times until the fracture occurs. The results indicate that the percentage of elongation of pre-strained specimens
with several times are smaller than that of the conventional uniaxial tensile test. The transition process from the
plane strain path to the uniaxial tensile strain path leading to the additional strain hardening is the main reason
for the reducing formability of sheet metal. While the percentage of elongation of specimens pulled through
drawbead inserts many times until the fracture occurs are evidently higher than that of the conventional uniaxial
tensile test and could reach close to 100%. Bending and unbending under tension has a significant effect on the
increased percentage of elongation.
1. Introduction pieces is the A-side of Audi A6 fender. When the test piece was stamped
especially during the process of passing through the drawbead set, the
The formability of sheet metal is the capability which undergoes major strains on the A surface of the piece were bigger than areas where
plastic deformation to a given shape without defects (Banabic et al., the material has not passed through drawbead, such as the flatter part
2010). The most commonly used tool in assessing the formability of of the test piece shown in Fig. 2. The final thinning in thickness di-
sheet metal is the forming limit diagram (FLD) developed by Keeler and rection in that area where the sheet metal flows through drawbead
Backofen (1963) and Goodwin (1968), which is composed of the inserts is close to 50%, which means the thinning in thickness direction
maximum values of the principal strains determined by measuring the is very serious, but the sheet metal is very safe in practice. By numerical
strains at failure, widely used in factories and research laboratories simulating the forming process of the sheet metal, the forming limit
(Stoughton and Zhu, 2004). The forming limit curve (FLC) is the line strains of that area is also very high, above the normal FLC.
consisting of limit principal strains separating the FLD into two regions. As we know, the FLC is only valid under following conditions: (1)
The deformed sheet metal is defined as failure if strain combinations lie straight or proportional strain path; (2) absence of bending; (3) without
above the FLC, otherwise the deformed sheet metal is treated as safe through-thickness shear; (4) no normal stress or through-thickness
(Nurcheshmeh and Green, 2016), as shown in Fig. 1. stress (Emmens and van den Boogaard, 2009). When the conditions
However, a phenomenon that when the sheet metal flows through above are not satisfied, the FLC is not always effective (Stoughton,
drawbead, the limit strains of sheet metal lie above the forming limit 2000).
diagram indicating fracture in theory, while it is safe in practice, vio- In terms of strain paths, Laukonis and Ghosh (1978) conducted an
lates the normal forming limit curve has been found after testing a large experiment on aluminum-killed steel and 2036-T4 aluminum by pre-
number of automobile panels. The test was done by stamping engineers straining in balanced biaxial tension. The FLC of aluminum-killed steel
in Faw Mould Manufacturing co.LTD. The mostly used material of the decreased while the position of FLC of 2036-T4 aluminum just moved
test is BH180 and the percentage of elongation is 29%. One of the test horizontally with the pre-strain increasing. Graf and Hosford (1993)
⁎
Corresponding author.
E-mail address: zhangzb@hust.edu.cn (Z. Zhang).
https://doi.org/10.1016/j.jmatprotec.2017.11.051
Received 2 July 2017; Received in revised form 22 November 2017; Accepted 24 November 2017
Available online 26 November 2017
0924-0136/ © 2017 Elsevier B.V. All rights reserved.
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
punch curvature on the stretching limit of AKQD steel and the limit
strains increased with the increasing punch curvature at constant ma-
terial thickness. Tharrett and Stoughton (2003) conducted a series of
stretch bending tests with various curvature punches and the limit
strains of small radius were larger and the position of FLC was higher.
Kitting et al. (2009) determined the forming limit of H340LAD, CP800,
and DP800 under a range of punch radius and deduced that bending
had an influence on the forming limit. With the decreasing of punch
radius, the forming limit became higher, especially when the punch
radius was below 10 mm. Emmens and van den Boogaard (2008) in-
troduced the bending into the conventional tensile test on DC04 and
DC06 with different thicknesses and discovered the percentage of
elongation was increased greatly. Fictorie et al. (2010) conducted the
Nakazima test with four kinds of radii of punches on DC06 and AA5051,
and the forming limit increased with decreasing of radius of punch,
especially in the plane strain region. The effect of bending on forming
limit of sheet metal cannot be neglected.
Fig. 1. Schematic diagram of the forming limit curve. In addition, the forming limit can be increased if the through-
thickness stress or through-thickness shear involves. Smith et al. (2003)
put forward a new sheet metal formability model considering the
carried out an experiment on Al 2008-T4 pre-strained in uniaxial,
through-thickness stress and proved that the formability increased with
biaxial and plane-strain tension parallel and perpendicular to the
the through-thickness stress. Assempour et al. (2010) predicted the
rolling direction and demonstrated that pre-straining in biaxial tension
forming limit diagram taking into account the normal stress and ver-
lowered the entire FLC, while pre-straining in uniaxial tension raised
ified that the FLD shifted up when the normal stress increased. Zhang
the right side of FLC. Pre-straining in plane strain tension raised both
et al. (2014) established constitutive models with M-K methods and
sides away from the minimum of the FLC. Then Graf and Hosford
different yield criterions to predict the effect of through-thickness stress
(1994) continued an experiment on Al 6111-T4 and found that if the
on the FLC and the FLC enhanced with increased through-thickness
directions of the principal strains were rotated, pre-straining in uniaxial
stress. Allwood and Shouler (2009) proposed a new generalized
or plane strain tension lowered the FLC besides the aforementioned.
forming limit diagram allowing for the normal stress and through-
That is to say, changing strain paths has an influence on the formability
thickness shear stress and demonstrated that the forming limits can be
of sheet metal.
increased significantly by both normal compressive stress and through-
The most commonly used method in experiment determination of
thickness shear. What’s more, Eyckens et al. (2009) established a model
the FLD is Nakazima test (out-of-plane) and Marciniak test (in-plane),
considering the effect of through-thickness shear with M-K method by
acquiring strains from uniaxial tensile test to plane strain and then to
investigating the incremental forming, and the results showed that
biaxial tension by testing specimens with different widths (Panich et al.,
formability was increased for all in-plane strain modes if through-
2013). However, the bending effect is ignored to meet the condition of
thickness shear was present.
membrane forces in the tests. Ghosh and Hecker (1974) compared the
On the other hand, the drawbead has been widely used in sheet
in-plane stretching with the out-of-plane stretching and found that the
metal forming processes, especially in automotive industry and elec-
out of plane stretching produced larger limiting strains under identical
trical appliances (Kim et al., 1997). By setting appropriate geometric
degree of biaxial tension. Charpentier (1975) investigated the effect of
Fig. 2. The phenomenon that when the sheet metal flows through drawbead, the forming limit strains of the sheet metal are above the FLC, which illustrates serious fracture in theory,
while it is very safe in practice. (a) the test piece (A-side of Audi A6 fender) in practice. (b) the forming limit strains of the test piece are above the normal forming limit curve. (c) the
simulation result of the stamping process of the test piece (A-side of Audi A6 fender).
284
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Table 1
Chemical composition of WDQ steel by weight percentages.
C Si Mn P S Cu ALT ALS
Ni Cr Ti Mo V Nb N B
0.006 0.01 0.0677 0.001 0.001 0.003 0.0029 0.0032
Table 2
Fig. 3. Schematic diagram of the sheet metal flowing through semicircle drawbead in- Mechanical properties of WDQ steel.
serts. Symbol meaning: R: radius of male bead, H: height of male bead, r1: entrance radius
of contra bead, r2: exit radius of contra bead. Number 0−1 0−2 45−1 45−2 90−1 90−2
285
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
drawbead is the most widely used drawbead type, shown in Table 3. percentage of elongation were obtained.
The length of the specimen is designed 250 mm, with about 50 mm
griped by the clamping unit and the rest pulled through drawbead in- 2.4. Drawbead test (two groups of drawbead inserts selected, passing
serts, and the width of specimen is designed as 30 mm, as the width of through many times until the fracture occurs)
drawbead groove is only 70 mm.
In order to measure the strains of specimens flowing through var- The test method was referring to the description of Section 2.2. The
ious drawbead inserts precisely, 2.5 mm square grids were printed on variations were two folds. One was that two groups were selected from
one surface of the specimen using electrochemical corrosion method. the nine groups of drawbead inserts. Another was that the specimen has
Before that, the specimens were degreased and cleaned with alcohol. to go through drawbead inserts many times until the fracture occurs
The Auto Grid Com-smart device was used to take pictures of deformed rather than several times.
grids on the surface of the specimen when the specimen has been pulled Two problems occurred when the test was done. One was that the
through drawbead inserts, and the length stain, width strain and grids on the surface of the specimens suffered from seriously abrasion
thickness strain were measured by software analysis. The principal of after many times, leading to the difficulty in measuring the strains by
the Auto-Grid Com-smart is the constant volume assumption. Auto Grid Com-smart, as shown in Fig. 6(a). Another was that the
In general, one group of drawbead inserts was put into the grooves clamped part of the specimen ruptured limiting the further test, as
of the black box and one specimen with one surface printed 2.5 mm shown in Fig. 6(b).
square grids was gripped by the clamping unit of drawbead test ma- To measure the length strains of specimens which suffered from
chine. Adjusting the clamping unit down to the location about seriously abrasion, we measured the length of 30 grids, 40 grids and 50
30 ± 3 mm to the upper surface of the black box in Fig. 5(a) and grids manually, and then the final length strain was obtained by aver-
setting the gap between the male bead and the contra bead close to 1.1 t aging them. Moreover, the zone selected must be very careful because
to make sure the specimen was under bending. Setting the pulling speed the length strain is not uniform which will be described in the following
2.1 mm/s, and then pulling the specimen through drawbead inserts by part. As shown in Fig. 6(c), the selected zone was almost in the middle
driving the hydraulic equipment. When the specimen was pulled part of the deformed specimens. In order to solve the fracture of
through drawbead inserts, the grids on the surface of the specimen were clamped part of the specimen, about 5 mm or 10 mm were adjusted to
deformed. The strains of the specimens were acquired by using the Auto the normal position by moving up or down every time when the spe-
Grid Com-smart device. By changing the groups of drawbead inserts cimen was clamped to make sure the clamping part of the specimen is
and adjusting the number of times of the specimens passing through, not always at the same place.
different pre-strained specimens were obtained.
3. Results and discussion
2.3. Pre-strained specimens + Conventional uniaxial tensile test
3.1. Drawbead test (one time, two times and three times)
The pre-strained specimens above were wire-electrode cut into the
uniaxial tensile test samples. Conventional uniaxial tensile test was The strains of specimens were measured by Auto Grid Com-smart,
employed again on Zwick Z050 electronic tensile testing machine. Basic including the length strain, width strain and thickness strain, as shown
mechanical properties, including yield strength, tensile strength, and in Fig. 7. It depicts that the width strain is almost unchanged during the
286
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Table 3 specimen going on, the resistance force tends to be relative stable,
Geometrical parameters of nine groups semicircle drawbead inserts. leading to the length strain and the thickness strain decreasing in some
degree. However, the length strain and the thickness strain are not
Name Male bead Contra bead
uniform in stable area because the lubrication is not uniform in the
Radius Height Entrance radius Exit radius operation resulting in the resistance force fluctuating.
To contrast the pre-strains caused by various drawbead inserts and
Symbol R H r1 r2 the number of times. Here we chose the average length strain of rela-
Value 6 4, 5, 6 1.5, 3, 4.5 1.5, 3, 4.5
tively stable area of the specimen as a criterion. The relationship be-
Note: (1) the units of radius, height, entrance radius and exit radius are mm; (2) the tween the pre-strains in length direction and the drawbead inserts
entrance radius and exit radius are the same of one group drawbead insert. (3) nine containing three groups of heights of male bead and three groups of
combinations (three groups of Heights and three groups of Entrance radius and Exit ra- radii of entrance and exit of contra bead and the number of times when
dius, namely, 3 × 3) are studied. (4) three repeat experiments are done for each test. the specimen was pulled through is shown in Fig. 8. It shows that with
the increase of height from 4 mm to 6 mm, the pre-strains are in-
process when the specimen is pulled through drawbead inserts, and creasing with no relation to the radius and the number of times. Under
compared with the length strain and thickness strain, the width strain is the conditions of same entrance and exit of contra bead and the number
very small, close to zero, which means the process of specimen flowing of times, the increase of height of male bead means the specimen has to
through drawbead inserts can be considered as plane strain deforma- go through drawbead inserts with more resistance forces causing more
tion. Moreover, in the beginning, the length strain and the thickness deformations. In the meantime, with the increase of radius, the pre-
strain increase fast as the resistance force starts to increase and with the strains decrease with no relation to the height and the number of times.
287
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
288
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Fig. 10. The relationship between the tensile strength of pre-strained specimens after
uniaxial tensile test and the effective pre-strains of specimens after flowing through
various drawbead inserts with one time, two times and three times.
Fig. 11. The relationship between the percentage of elongation of pre-strained specimens
after uniaxial tensile test and the effective pre-strains of specimens after flowing through
various drawbead inserts with one time, two times and three times.
Fig. 8. The relationship between the pre-strains in length direction and the drawbead
inserts containing three groups of heights of male bead and three groups of radii of en-
trance and exit of contra bead and the number of times of the specimens which were
pulled through. (a) one time. (b) two times. (c) three times.
Fig. 12. The comparison of engineering stress-engineering strain curve between pre-
strained specimens flowing through various drawbead inserts and different number of
times and the conventional uniaxial tensile curve.
first strain path, namely, the plane-strain pre-strain, then the following
uniaxial tensile test as the second strain path, the results here shows
that the formability of sheet metal has reduced. As described in the
introduction, Graf and Hosford conducted a test on the Al 6111-T4 and
Al 2008-T4 to investigate the effect of changing paths on forming limits,
and they found that the plane-strain pre-strain raised both sides of the
FLC away from the minimum, which means the plane-strain pre-strain
Fig. 9. The relationship between the yield strength of pre-strained specimens after uni- raises the formability of sheet metal, as shown in Fig. 13.
axial tensile test and the effective pre-strains of specimens after flowing through various However, the results here show that the left side of the FLC falls
drawbead inserts with one time, two times and three times. down, contrary to the results of Graf and Hosford. The reasons may be
that, in the first place, the test material is different. Graf and Hosford
used the aluminum sheets, but the authors used the steel sheets. More
289
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Fig. 13. The forming limit strains of Al 6111-T4 after prestrained near plane-strain ten- Fig. 15. The relationship between the times of specimens after flowing through two
sion (Graf and Hosford, 1994). groups of drawbead inserts and the percentage of elongation of specimens.
important is that the process of pre-strain is different evidently. The pulled through. So we choose this drawbead combination. While the
plane-strain pre-strain is obtained by stretching wide sheet under the pre-strain of drawbead combination (the height is 5 mm and the en-
uniaxial tensile testing machine in Graf and Hosford’s test, while it is trance radius and exit radius are 3 mm) is in the middle of all pre-strain
acquired by pulling the specimen through the drawbead inserts in this values among all the drawbead combinations. We choose this drawbead
test. It is quite clear that the process of pulling through drawbead is combination as a contrast. Here we define the drawbead inserts with
different from that of stretching wide sheet under the uniaxial tensile the height of 6 mm and the radius of entrance and exit of 1.5 mm as
testing machine, which involves bending and unbending under tension. group A and the drawbead inserts with the height of 5 mm and radius of
In order to explain more clearly the effect of passing through entrance and exit of 3 mm as group B.
drawbead on the formability of specimens. We construct an effective The relationship between the percentage of elongation of specimens
stress − effective strain curve based on the uniaxial tensile curves of and the number of times of specimens pulling through two groups of
pre-strained specimens, as shown in Fig. 14. From the figure, we can see drawbead inserts is shown in Fig. 15. It depicts that with the increase of
that the process of passing through drawbead makes the specimen the numbers of times, the percentage of elongation of specimens with
undergoes additional hardening compared to the single uniaxial tensile two drawbead inserts also increases almost linear. The trend of the
strain path. linear lines is that the resistance force is relatively invariable for any
Then, if we suppose the instability condition: group of drawbead inserts, when the pulling speed from the clamping
unit and the gap between the male bead and contra bead are fixed.
dσ/dε = σ
While the increment of the length strain is very related to the resistance
known as the considère condition, premature necking is certain to take force. When the specimen was pulled through the drawebead inserts
place for the pre-strained specimens. Therefore, the transition process one time, under the same resistance force, the increment of the length
from the plane-strain path to the uniaxial tensile strain path leading to strain, or the percentage of elongation, is relatively fixed. Moreover, the
the additional strain hardening is the main reason for the reducing gradient of lines of group A is higher than that of group B, as the pre-
formability of sheet metal. strain of group A is bigger than the group B in Section 3.1 and the
increment of the length strain of group A is more than that of group B.
Therefore, we can deduce that the gradient between the lines propor-
3.3. Drawbead test (many times until the fracture occurs)
tional to the change in pre-strain caused by different drawbead inserts.
Here we briefly explain the maximum number of times. In Section
As described in Section 2.4, two groups of drawbead inserts were
2.4, two problems exist in the drawbead test (many times): one is the
selected from nine groups. The intention of the retested samples se-
measure of length strains when the surface grids on the sheet suffer
lected was based on the pre-strain values which have been shown in
from abrasion. Another is the clamping part of the specimen ruptures
Section 3.1 Drawbead test. As the pre-strain of drawbead combination
unexpectedly. Although the method of acquiring the length strains by
(the height is 6 mm and the entrance radius and exit radius are 1.5 mm)
multiple measuring as relatively accurate as possible is adopted, the
is the biggest under the same number of times when the samples were
number of times of group B is still restricted to 13 with almost no grids
recognized. Meanwhile, albeit we tried to adjust the distance of the
clamping part to keep the specimen not clamped in the same zone every
time, the result is not good, leading to the number of times of group A
only reaching 10.
However, ignoring the two problems, the percentage of elongation
of group A is still much larger than that of the conventional uniaxial
tensile test and reaches close to 100%. Then we have to ask why the
percentage of elongation can be so large when the specimen was pulled
through drawbead inserts than that in the conventional uniaxial tensile
test and why the percentage of elongation could be so large reaching
nearly 100%.
As described in the introduction, the process of passing through
drawbead can be seen as the cycle of bending, unloading and un-
bending with tensile forces. Here we only consider the effect of bending
Fig. 14. The comparison of effective stress − effective strain between the constructed and unbending on the deformation of the specimen.
plane strain curve based on the uniaxial tensile curve after plane strain pre-strain and the During the process of passing through one round corner (the
conventional uniaxial tensile curve.
290
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Fig. 16. The schematic of pure bending deformation. Left: the specimen
with thickness t and length L0. Right: the bending moment M with radius
R.
Fig. 17. The strain distribution and stress distribution of the sheet under
pure bending.
Entrance corner or the Exit corner or the Height corner) of the spe- At any time, the tension force T per unit width and bending moment M
cimen, we can consider it as the process of bending under tension. A per unit width of the distance y from the neutral axial was calculated as
situation of rigid, perfectly plastic material with plane strain flow stress following:
S was assumed. For a sheet of thickness t, and length L0 that is bent to a
radius R, at the same time, a tension force per unit width T and a t /2
works, namely the pure bending situation, as shown in Fig. 16. The
strain distribution and stress distribution were presented in Fig. 17.
t /2
eb = y/R (neglect the second order effects)
And the stress is
M= ∫ σx × y × 1 × dy
−t /2
S y>0
σx = ⎧ When the β ≤ 1, the neutral layer has moved the distance u, and
⎨ −S y<0
⎩
1 1
Then when the tension force T increases, the neutral axis moves inside, u = (ea/ eb) × × t = βt
2 2
as shown in Fig. 18 and the strain distribution and the stress distribu-
tion were shown in Fig. 19.
At this time, the strain of the distance y from the neutral layer is −u t /2 −u t /2
e = ea + eb = (β + 1)eb = (β + 1)y/R (neglect the second order ef- T= ∫ σx × 1 × dy + ∫ σx × 1 × dy = ∫ −Sdy + ∫ Sdy=2Su
fects) −t /2 −u −t /2 −u
Fig. 18. The schematic of bending and tension deformation. Left: the
specimen with thickness t and length L0. Right: the bending moment M
and tension T with radius R.
291
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Fig. 19. The strain distribution and stress distribution of the sheet under
bending and tension.
Fig. 20. Left: the relationship between the tension force T and the strain
ratio β. Right: the relationship between the moment M and the strain ratio
β.
292
J. Ke et al. Journal of Materials Processing Tech. 254 (2018) 283–293
Fig. 21. Multiple necks developed along the specimen after passing
through the drawbead many times.
bead and the number of times when the specimens were pulled Emmens, W.C., van den Boogaard, A.H., 2008. Extended tensile testing with simultaneous
through drawbead inserts have an influence on the pre-strains of bending. Int. Deep Drawing Res. Group 1–12.
Emmens, W.C., van den Boogaard, A.H., 2009. An overview of stabilizing deformation
sheet metal. In briefly, the higher the height of male bead, the mechanisms in incremental sheet forming. J. Mater. Process. Technol. 209,
smaller the entrance and exit radii of contra bead, the more number 3688–3695.
of times, the more pre-strains can be obtained. Meanwhile, ad- Emmens, W.C., van den Boogaard, A.H., 2011. Cyclic stretch-bending: mechanics, sta-
bility and formability. J. Mater. Process. Technol. 211, 1965–1981.
justing the radius of entrance and exit of contra bead is more effi- Eyckens, P., Van Bael, A., Van Houtte, P., 2009. Marciniak–Kuczynski type modelling of
cient than the height of male bead. the effect of Through-Thickness Shear on the forming limits of sheet metal. Int. J.
(2) The yield strength and the tensile strength increase with the in- Plast. 25, 2249–2268.
Fictorie, E., van den Boogaard, A.H., Atzema, E.H., 2010. Influence of punch radius in a
crease of pre-strains, while the percentage of elongation decreases nakazima test for mild steel and aluminium. Int. J. Mater. Form. 3, 1179–1182.
with the increase of pre-strains. Under the condition of same pre- Firat, M., 2007. Computer aided analysis and design of sheet metal forming processes.
strain, the yield strength, tensile strength and percentage of elon- Mater. Des. 28, 1311–1320.
Ghoo, B., Keum, Y., 2000. Expert drawbead models for sectional FEM analysis of sheet
gation are not the same, which relates to the type of drawbead
metal forming processes. J. Mater. Process. Technol. 105, 7–16.
inserts. Moreover, the percentage of elongation of pre-strained Ghosh, A.K., Hecker, S.S., 1974. Stretching limits in sheet metals?In-plane versus out-of-
specimens were evidently smaller than that of the conventional plane deformation. Metall. Trans. A 6A, 1665–1669.
uniaxial tensile test. The transition process from the plane-strain Goodwin, G., 1968. Application of Strain Analysis to Sheet Metal Forming Problems in the
Press Shop. SAE Technical Paper.
path to the uniaxial tensile strain path leading to the additional Graf, A., Hosford, W., 1994. The influence of changing strain path on forming limit
strain hardening is the main reason for the reducing formability of diagrams of Al 6111-T4. Int. J. Mech. Sci. 36, 897–910.
sheet metal. Graf, A., Hosford, W., 1993. Effect of changing strain path on forming limit diagrams of Al
2008-T4. Metall. Trans. A 24A, 2503–2512.
(3) When the specimens were pulling through drawbead inserts many Keeler, S., Backofen, W., 1963. Plastic instability and fracture in sheets stretched over
times until the fracture occurs, the percentage of elongation exceeds rigid punches. ASM Trans. Q 56, 25–48.
that of the conventional uniaxial tensile test and reaches close to Kim, C., Im, Y., Heo, Y., Kim, N., Jun, G., Seo, D., 1997. Finite-element analysis and
experimental verification for drawbead drawing processes. J. Mater. Process.
100%. Bending and unbending under tension has a significant effect Technol. 72, 188–194.
on the increased percentage of elongation. Kitting, D., Koplenig, M., Ofenheimer, A., Pauli, H., Till, E.T., 2009. Application of a
concave-side rule approach for assessing formability of stretch-bent steel sheets. Int.
J. Mater. Form. 2, 427–430.
Acknowledgements Laukonis, J.V., Ghosh, A.K., 1978. Effects of strain path changes on the formability of
sheet metals. Metall. Trans. A 9A, 1849–1956.
We express our sincere appreciation to the Research and Nurcheshmeh, M., Green, D.E., 2016. Prediction of forming limit curves for nonlinear
loading paths using quadratic and non-quadratic yield criteria and variable im-
Development center of Wuhan iron and steel (group) corp for providing
perfection factor. Mater. Des. 91, 248–255.
the equipment and sheet metals. This research did not receive any Panich, S., Barlat, F., Uthaisangsuk, V., Suranuntchai, S., Jirathearanat, S., 2013.
specific grant from funding agencies in the public, commercial, or not- Experimental and theoretical formability analysis using strain and stress based
for-profit sectors. forming limit diagram for advanced high strength steels. Mater. Des. 51, 756–766.
Samuel, M., 2002. Influence of drawbead geometry on sheet metal forming. J. Mater.
Process. Technol. 122, 94–103.
References Smith, L.M., Averill, R.C., Lucas, J.P., Stoughton, T.B., M, P.H., 2003. Influence of
transverse normal stress on sheet metal formability. Int. J. Plast. 19, 1567–1583.
Stoughton, T.B., Zhu, X., 2004. Review of theoretical models of the strain-based FLD and
Allwood, J.M., Shouler, D.R., 2009. Generalised forming limit diagrams showing in- their relevance to the stress-based FLD. Int. J. Plast. 20, 1463–1486.
creased forming limits with non-planar stress states. Int. J. Plast. 25, 1207–1230. Stoughton, T.B., 2000. A general forming limit criterion for sheet metal forming. Int. J.
Assempour, A., Nejadkhaki, H.K., Hashemi, R., 2010. Forming limit diagrams with the Mech. Sci. 42, 1–27.
existence of through-thickness normal stress. Comput. Mater. Sci 48, 504–508. Sun, G., Li, G., Gong, Z., Cui, X., Yang, X., Li, Q., 2010. Multiobjective robust optimization
Banabic, D., Barlat, F., Cazacu, O., Kuwabara, T., 2010. Advances in anisotropy and method for drawbead design in sheet metal forming. Mater. Des. 31, 1917–1929.
formability. Int. J. Mater. Form. 3, 165–189. Tharrett, M., Stoughton, T.B., 2003. Stretch-bend Forming Limits of 1008 AK Steel. SAE
Charpentier, P.L., 1975. Influence of punch curvature on the stretching limits of sheet Technical Paper.
metal. Metall. Trans. 5, 2161–2164. Zhang, F., Chen, J., Chen, J., 2014. Effect of through-thickness normal stress on forming
Courvoisier, L., Martiny, M., Ferrom, G., 2003. Analytical modelling of drawbeads in limits under Yld2003 yield criterion and M-K model. Int. J. Mech. Sci. 89, 92–100.
sheet metal forming. J. Mater. Process. Technol. 133, 359–370.
293