Proceedings of the Institution of Mechanical

Engineers, Part B: Journal of Engineering

Manufacture
http://pib.sagepub.com/

Numerical simulation of sheet stamping process using flexible punch

M Ramezani, Z M Ripin and R Ahmad
Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture 2009 223: 829
DOI: 10.1243/09544054JEM1453

The online version of this article can be found at:

http://pib.sagepub.com/content/223/7/829

Published by:

http://www.sagepublications.com

On behalf of:

Institution of Mechanical Engineers

Additional services and information for Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering
Manufacture can be found at:

Email Alerts: http://pib.sagepub.com/cgi/alerts

Subscriptions: http://pib.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://pib.sagepub.com/content/223/7/829.refs.html

>> Version of Record - Jul 1, 2009

What is This?

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

 829

Numerical simulation of sheet stamping

process using flexible punch
M Ramezani*, Z M Ripin, and R Ahmad
School of Mechanical Engineering, University Science Malaysia, Penang, Malaysia

The manuscript was received on 18 December 2008 and was accepted after revision for publication on 11 March 2009.

DOI: 10.1243/09544054JEM1453

Abstract: Different from the conventional stamping process, rubber forming uses a rubber pad
contained in a rigid box to act as a punch. The main attraction of the process is its simplicity, and it
can be used for prototype development or for low-volume production. In this study, an axisym-
metric rubber-pad forming operation is investigated via numerical simulations and experiments.
Some key process parameters, such as rubber material, hardness of flexible punch, stamping
velocity, rubber-pad thickness, and friction are studied in detail. Good correlation is achieved
between the finite element predicted and experimentally measured thickness–thinning in the
formed part. Stress and strain distributions in the specimen are investigated numerically during
the process using ABAQUS/Standard.

Keywords: finite element, flexible punch stamping, rubber-pad forming

1 INTRODUCTION flexible tools instead of conventional metallic tools

are that:
Stamping is a metal forming process by which sheet
(a) the flexible pad can be used for several different
metal workpieces are punched using a press tool that
shapes of workpiece;
is loaded on a machine press or stamping press to
(b) the alignment and mismatch problems are
form the sheet into a desired shape. The conventional
eliminated;
stamping process is performed through a rigid
(c) lubrication is usually not needed;
punch, which, together with a blank holder, forces
(d) parts with excellent surface finish can be formed
the sheet metal to slide into a die and comply with
with no tool marks.
the shape of the die itself.
The rubber forming process uses a rubber pad However the rubber-pad forming process has cer-
contained in a rigid box in which one of the tools is tain disadvantages, such as:
replaced by the rubber pad. The rubber pad in this
(a) a higher capacity press is usually required;
method may be constructed either solidly or lami-
(b) a lack of sufficient forming pressure results in
nated. The laminated pad is comprised of sheets of
parts with less sharpness that may require sub-
rubber placed over one another. The advantage of
sequent hand works;
this is that the working surface can be restored by
(c) the lifetime of the flexible pad is limited [1].
turning the top layer over or replacing it.
In the aircraft industry most of the sheet metal Several studies have been carried out to analyse
parts, such as frames, seat parts, ribs, windows, and rubber-pad forming. Browne and Battikha [2] pre-
doors, are fabricated using the rubber-pad forming sented an experimental study of the rubber-pad
processes. In other industries, for instance the auto- forming process to investigate the capability of the
motive industry, this process is mainly used for pro- process and to optimize the process parameters.
totypes or pilot productions. The advantages of using They analysed the use of different types of lubricants
at the blank and its interfaces. The dependence of the
*Corresponding author: School of Mechanical Engineering, clamping force for the prevention of wrinkling and
University Science Malaysia, Nibong Tebal, Penang 14300, cracking on the type of blank material was also inves-
Malaysia. tigated. Thiruvarudchelvan [3] and Thiruvarudchelvan
email: [email protected] and Tan [4] introduced several techniques for the

JEM1453 Ó IMechE 2009 Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

830 M Ramezani, Z M Ripin, and R Ahmad

use of flexible tools in metal forming. They presented

principles involved in friction-aided sheet metal
forming techniques, the design of the flexible tools,
the actual prototype devices fabricated and tested,
and the experimental data from forming operations.
Sala [5] optimized the rubber-pad forming process
of an aluminium alloy fuselage frame belonging to
the AerMacchi MB-339 trainer aircraft using a spe-
cific finite element code. Several effects, depending
on stamping velocity, component geometry, sheet
metal heat treatment, elastomeric rubber-pad con-
stitutive law, and thickness were taken into account.
Dirikolu and Akdemir [6] carried out a three-dimen-
sional, finite-element (FE) simulation study con-
cerning the flexible forming process, to investigate
the influence of rubber hardness and blank material
type on stress distribution in the formed blank.
Ramezani et al. [7] presented theoretical models for
Fig. 1 Schematic of the rubber-pad forming process
static and kinetic frictions in rubber–metal and
metal–metal contacts. They applied the new friction
models to FE simulation of an axisymmetric rubber-
pad forming process. The results of their analysis
clearly showed that the new friction models provided
better agreement between experiments and results of
numerical simulations. Peng et al. [8] investigated the
sheet soft-punch stamping process to fabricated
microchannels via numerical simulations and
experiments. Grain size of the sheet metal, hardness
of the soft punch, and lubricant condition were stu-
died and detailed in their paper. However, none of
these papers studied the influence of using different
types of rubber on the process. A detailed study of
stress and strain analysis in the sheet, and the
mechanism of deformation during process are also
missing in these works. These weaknesses necessitate
a new approach on rubber-pad forming, which con-
sider the influence of different parameters on the Fig. 2 Experimental set-up and formed blanks
process.
In this paper, the rubber-pad forming process is
studied by numerical simulation and experiments. blank of 40 mm diameter and 1 mm thick. Before
Some key process parameters such as rubber mate- each test, the die was cleaned and lubricated with
rial, hardness of flexible punch, stamping velocity, mineral oil in order to minimize friction between the
rubber-pad thickness, and friction conditions are die and the blank. A servo-hydraulic universal testing
investigated in detail. Non-linear FE analysis using machine (ESH300) with maximum press force capa-
commercial software ABAQUS/Standard is con- city of 300 kN was used for the rubber-pad forming
ducted to analyse stress and strain distribution and process. The process begins with the die placed on
deformation mechanisms during the rubber-pad the hydraulic testing machine. The aluminium blank
forming process. is then introduced between the die and the flexible
punch. At this stage, the flexible punch moves down
to stamp the blank. The experimental set up and the
2 EXPERIMENTS AND FE MODELLING formed blanks are shown in Fig. 2.
In order to simulate the rubber-pad forming, a finite
A schematic of the arrangement used for the experi- element model is built in commercial software ABA-
mental procedure is shown in Fig. 1. The die and QUS/Standard according to the geometry of the
rubber pad retainer are made of mild steel. Natural experimental apparatus as described above. A quasi-
rubber and silicon rubber of Shore Hardness 50A and static non-linear approach with negligible temperature
diameter 40 mm are used to stamp an aluminium effects is assumed for simulations. The non-linear

Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture JEM1453 Ó IMechE 2009

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

 Numerical simulation of sheet stamping process using flexible punch 831

Fig. 3 Experimental tensile stress–strain curve for an aluminium blank

approach is chosen because of the sheet plasticity and density and 50 Shore A hardness and v ¼ 0.49997,
rubber hyperelasticity during simulations that have E ¼ 7.053 MPa are shown in Fig. 4.
highly non-linear behaviours. By taking advantage of An elasto–plastic material model with von-Mises
axisymmetry it is possible to simulate the die, blank, yield criterion is used for the blank. It is defined by
and rubber assembly as a two-dimensional axisym- giving the value of the uniaxial yield stress as a
metric model. The flexible rubber is modelled using function of uniaxial equivalent plastic strain. Flexible
CAX4RH elements. CAX4RH is a four-node bi-linear materials have non-linear stress–strain character-
axisymmetric quadrilateral, hybrid, constant pressure, istics for relatively large deformations. Under such
reduced integration, hourglass control element. The conditions, they are generally assumed as nearly
blank is modelled using a four-node bi-linear axisym- incompressible [10]. To model these hyper-elastic
metric quadrilateral, reduced integration, hourglass materials through finite element method (FEM), a
control CAX4R element. The interfaces between the die constitutive law based on total strain energy density
and the blank, and between the rubber and the blank, W has to be adopted. Among several approaches,
are modelled using an automatic surface-to-surface Mooney–Rivilin theory [11] is used based on the
contact algorithm. A Coulomb friction law is assumed polynomial development of total strain energy. The
and a coefficient of friction of 0.15 is assigned between form of the Mooney–Rivlin strain energy potential is
the die and the blank for lubricated condition. A coef- @W
ficient of friction of 0.25 is used between the rubber sij ¼ ð1Þ
@"ij
and the blank (for more details about friction model-
ling see Ramezani et al. [7, 9, 10]). The values of coef-
ficients of friction are based on historical data for X
n
1
similar cases (for example Dirikolu and Akdemir [6]). W ¼ Ckm ðI1
3Þk þ ðI2
3Þm þ kðI3
1Þ2
kþm¼1
2
The die is modelled as a rigid body. The material
properties for the aluminium blank (r ¼ 2780 kg/m3, ð2Þ
v ¼ 0.334, E ¼ 67.3GPa) were determined from a tensile
where W is the strain energy per unit of reference
test performed on the blank material according to
volume; I1, I2, I3 are the strain invariants; k is the bulk
ASTM E 8M standard (see Fig. 3). Compression tests modulus; I3 ¼ 1 for incompressible material beha-
were also carried out according to ASTM D575 stan- viour; Ckm is the constant of the Mooney–Rivlin
dard to determine the material properties of the nat- material model. Usually two Mooney–Rivlin para-
ural rubber and silicon rubber. Both tensile and meters (C10 and C01) are used to describe hyper-
compression tests were carried out using model 4501 elastic rubber deformation. These parameters can be
INSTRON universal testing machine. The results of the determined by experiments. The Mooney–Rivlin
compression test of natural rubber having 50 Shore A material model has previously been used with
hardness with r ¼ 1200 kg/m3, v ¼ 0.49997, E ¼ 59.656 success to predict the behaviour of hyper-elastic
MPa and silicon rubber having 1250 kg/m3 specific materials (for example MacDonald and Hashmi [12]

JEM1453 Ó IMechE 2009 Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

832 M Ramezani, Z M Ripin, and R Ahmad

Fig. 4 Experimental compression stress–strain curve for natural and silicon rubber

Table 1 Mechanical properties of flexible punches

Hardness Mooney–Rivlin Mooney–Rivlin

Rubber type Shore A ( ) constant C10 (MPa) constant C01 (MPa) E (MPa) Poisson’s ratio

Natural rubber 50 2.307 0.157 59.656 0.4997

Silicon rubber 50 0.394 0.092 7.053 0.4997

and Girard et al. [13]). The Mooney–Rivlin constants in the second and last stage, it seems that cracks are
evaluated by ABAQUS using the compression test the main imperfection in this forming process.
data are listed in Table 1. Cracks occur when a blank undergoes stretching
Since the model is developed by taking advantage forces overcoming ultimate stress. During stretching,
of axisymmetry, the blank, die, and rubber nodes at the sheet metal first grows thin homogeneously and
the symmetry edges are restrained in the appropriate then shows localized necking, which can induce the
directions. The assembly is also constrained in the onset of cracks. In the current paper, the thinning in
Z-direction at the base in order to model the die base. thickness of the formed blank is used to evaluate
The forming loads are applied on the blank in terms the risk of sheet cracks during the forming process.
of displacements on the top surface of the rubber Figure 6 shows the simulated formed part and the
pad. The displacements are assigned to be equal to photo of experimental results. It is clear to see that
the displacements of the punch measured during the the simulation results match very well with the
experiments. experiments.

3.2 In-plane maximum principle stress

3 RESULTS AND ANALYSIS
Figure 7 shows symbol plots of in-plane maximum
principle stress in the workpiece with respect to the
3.1 Forming stages
punch stroke. Symbol plots allow the visualization of
Figure 5 shows the step-by-step forming process the magnitude and direction of stress tensor in the
using natural rubber. From the simulation results, it form of arrows superimposed on the model. The
is clear that the rubber-pad forming process contains length of the arrow indicates the magnitude of the
three different stages: the first is a flexible punch, tensor and the direction of the arrow indicates its
self-deformation stage; the second is the blank direction. According to Fig. 7, the material is forced
deformation under the press of flexible punch to gradually downward into the die as the punch stroke
reach the bottom of rigid die; and the last stage is the increases. Figure 7(a) shows the rubber self-defor-
filling of the die cavity with blank until it fits the rigid mation stage. At this stage the workpiece is slightly
die. Because of the stress concentration at the blank compressed by the punch and only in zone A (which

Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture JEM1453 Ó IMechE 2009

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

 Numerical simulation of sheet stamping process using flexible punch 833

Fig. 5 The three stages of the rubber-pad forming process

Fig. 6 (a) Specimen cross-section after cutting measured by Alicona microscope, (b) FE model of
rubber-pad forming. The numbers labelled are the locations for thickness measurement

Fig. 7 Symbol plots of in-plane maximum principle stress

JEM1453 Ó IMechE 2009 Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

834 M Ramezani, Z M Ripin, and R Ahmad

is in contact with the die) the sheet is under tensile occurs at point 1, this being where the metal has been
stress. It can be seen from Fig. 7(b) that the cupping stretched most. This is comparable with the work of
initially begins as a hump at the sheet. As the process Peng et al. [8] where the numerical analysis of micro-
continues the cupping grows and the blank begins to stamping process showed similar results. Using
fill the die cavity. At this stage zone B is still under natural rubber as a flexible punch, the maximum
compressive stress. As the punch stroke increases, thinning is 11.7 per cent at point 1. The maximum
the tensile stress at zone A increases resulting in thinning decreases to 10.6 per cent using silicon
thinning at zone A. Figure 7(c) shows the maximum rubber. In order to investigate the influence of
principle stress after the workpiece contacted the die forming speed on the thinning of the workpiece, the
bottom. At this stage zone C is under tensile stress test was repeated using different punch velocities of
and the blank is almost fully formed, and during the Vp ¼ 5, 25, 50, 75, 100 (mm/min) and measured the
remaining time of the simulation zone D of the blank thinning of the specimen using an Alicona micro-
moves toward the die side. At full load application scope. After measurement, no significant changes
(Fig. 7(d)), the blank is almost in complete contact were found in thinning using these five different
with the die wall. As can be seen from Fig. 7(d), all the velocities. The maximum thinning at point 1 using
zones are stretched, which result in the thinning of natural rubber decreases to 11.3 per cent at Vp ¼ 100
the blank. At zone D the compressive stress occurs on (mm/min), which is not remarkable.
the punch side and the tensile stress occurs on the
die side. As the punch stroke increases, the com- 3.4 Load-stroke curves
pressive and tensile stresses increase in zone D and
with the end of the punch stroke, the workpiece Figure 9 shows the punch load–stroke curves, using
moves downward to fully contact the die side. natural rubber and silicon rubber as flexible punches,
obtained from experiments. As can be seen from the
curves, there are significant differences between the
3.3 Thickness distribution
punch loads necessary to form the blank using these
After forming the blank, the resultant geometry was two types of rubber. The maximum load at stroke
cut longitudinally using Struers metallurgical speci- 2.85 mm for natural rubber is 164.8 kN. The load
men cutter. Then, the thinning of the specimen is decreases to 67.3 kN using silicon rubber as a flexible
measured using an Alicona imaging infinite focus medium. Although silicon rubber has the advantage
microscope (IFM 2.1). The thickness is measured at 5 of a lower punch load, its lifetime is limited com-
points as determined in Fig. 6(a). The results of the pared to natural rubber and should be replaced after
measurements for the specimens formed at punch a few forming processes. The bending angle was
velocity Vp ¼ 5(mm/min) are shown in Fig. 8. As can measured at point B (see Fig. 6(b) ) of the formed
be seen from Fig. 8, the maximum thickness reduction part. The die angle at this point is 145 . An angle of

Fig. 8 Experimental thickness distribution of the formed part at Vp ¼ 5(mm/min)

Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture JEM1453 Ó IMechE 2009

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

 Numerical simulation of sheet stamping process using flexible punch 835

Fig. 9 Experimental load–stroke curves using natural and silicon rubber as flexible punches

Fig. 10 Comparison of experimental and FEM thinning results for a specimen formed by natural rubber

147 could be achieved using natural rubber and 151 materials, and loading conditions of the experiments,
using silicon rubber. It clearly shows that by using as explained earlier. Figure 10 shows the comparison
silicon rubber as flexible punch, sharp edges can not between experimental and FE results of the thinning
be achieved during forming, because of a lack of prediction using natural rubber. As can be seen from
sufficient forming pressure compared to natural Fig. 10, the thickness distribution is predicted accu-
rubber. According to the results, for a low pressing rately by the simulation. The maximum error of the
capacity and a low number of cycles, silicon rubber is FE prediction of thinning is 5 per cent for natural
the suitable choice for rubber-pad forming. For a rubber and 5.1 per cent for silicon rubber.
larger number of cycles and samples with sharp
edges, polyurethane and natural rubber are more 3.6 Effect of rubber Shore A hardness
suitable.
To investigate the influence of rubber Shore A hard-
ness on the process, urethane rubber with Shore
3.5 FE validation
A hardness 55 and 70 are adopted in a finite
In order to validate the results of the FE simulations, element analysis (FEA) simulation. The values of the
FE analysis was performed using the same geometry, Mooney–Rivlin constants for urethane (Table 2) are

JEM1453 Ó IMechE 2009 Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

836 M Ramezani, Z M Ripin, and R Ahmad

Table 2 Mechanical properties of polyurethane rubber

Hardness Mooney–Rivlin Mooney–Rivlin

Polyurethane rubber Shore A ( ) constant C10 (MPa) constant C01 (MPa) E (MPa) Poisson’s ratio

Type 1 55 0.382 0.096 2.868 0.4997

Type 2 70 0.736 0.184 5.517 0.4997

Fig. 11 Influence of Shore A hardness of urethane on thickness distribution of the formed part

obtained from the work of Dirikolu and Akdemir [6]. blank using different natural rubber-pad thicknesses.
The simulations revealed that as the Shore A hard- It can be seen from the figure that smaller rubber-pad
ness increases from 55 to 70, the maximum von- thickness causes slightly higher von-Mises stress in
Mises stress in the blank material increases from 257 the workpiece. The maximum von-Mises stress with
MPa to 261.5 MPa. Figure 11 shows the thickness rubber-pad thickness of 10 mm is 259.9 MPa. The
distribution of the parts stamped with various soft stress decreases to 258.8 MPa using rubber pad with a
punches. It can be seen from Fig. 11 that the thick- 20 mm thickness. This is in agreement with the pre-
ness curves are almost the same and Shore A hard- vious study of Sala [5]. The influence of rubber-pad
ness of rubber is not a decisive factor in the rubber- thickness on thickness thinning distribution is illu-
pad forming process. It shows that although the strated in Fig. 13. It can be observed that, as the
hardness is different, the capabilities to transfer the rubber-pad thickness decreases, the thinning of the
loading force are equal in close container. This is in formed blank increases. However, this increase in
agreement with the previous work of Dirikolu and thinning is not significant. The minimum thickness
Akdemir [6] where the numerical analysis of the of the workpiece at point 1 (see Fig. 6(a) ) is 0.844 mm
rubber-pad forming process showed the same using a 20 mm rubber pad and 0.835 mm using a
results. Comparing Fig. 11 and Fig. 10 indicates that 10 mm rubber pad.
the thinning of the blanks formed using urethane as
soft punch is quite the same as with natural rubber.
3.8 Effect of coefficient of friction
Friction condition is an important control parameter
3.7 Effect of rubber-pad thickness
that influences the formability and quality of the
The effect of rubber-pad thickness on the thickness formed parts. The limiting strain of a material is not
and stress distribution of the formed part is investi- directly changed by friction, but friction changes the
gated numerically. Three different rubber-pad thick- stress and strain distribution. The redistribution of
nesses, i.e. 10, 15, and 20 mm, are used in the stress and strain, can affect defects in metal forming,
simulations. Figure 12 shows the distribution of von- such as thinning [9]. Two friction interfaces are
Mises stress at the full load application in the formed defined in this study to investigate the effects of

Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture JEM1453 Ó IMechE 2009

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

 Numerical simulation of sheet stamping process using flexible punch 837

Fig. 12 Von-Mises stress distribution with different rubber-pad thicknesses

Fig. 13 Thickness distribution of formed parts with different rubber-pad thicknesses

friction. The friction coefficient between the flexible between the rubber and blank would help in pulling
punch and the metal sheet is f1 and the friction more material into the die and hence reduce thinning
coefficient between the metal sheet and the rigid die of the workpiece. According to Fig. 14, the higher
is f2. The value of f1 and f2 are defined from 0 (no coefficient of friction between blank and rigid die
friction) to 0.3. Figure 14 shows the numerical results results in more thinning of the blank. Higher friction
of thickness distribution of formed parts with differ- between blank and rigid die makes the metal sheet
ent coefficients of friction using natural rubber as a more difficult to flow into the die, resulting in more
flexible punch. From the figure, it can be seen that as stretching and thinning of the blank. The results
the friction between rubber and blank increases, the clearly show that providing the correct frictional
thickness thinning decreases. This is consistent with conditions between rubber–sheet and sheet–die
the finite element analysis of Peng et al. [8] where interfaces will improve the quality of specimen and
their results show similar effects. Higher friction avoid defects during the process.

JEM1453 Ó IMechE 2009 Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

838 M Ramezani, Z M Ripin, and R Ahmad

Fig. 14 Thickness distribution of formed parts with different coefficients of friction

Fig. 15 Stress distribution during the rubber-pad forming simulation with natural rubber

kept under control, as it can result in more material

3.9 Stress–strain analysis being pushed into the die, thus preventing excessive
thinning of the sheet. Axial stress is exactly the same
Figure 15 shows the development of von-Mises and
as radial stress early in the simulation but becomes
the principal axial and radial stresses in node A (see
Fig. 6(b) ) during the deformation process with nat- stable as the blank begins to form. Axial stress
ural rubber as a flexible punch. The cupping initially remains compressive for the entire simulation. At
develops owing to an increase in von-Mises stress t ¼ 30 s, the workpiece is now in full contact with the
and then is aided by an increase in radial stress. die and owing to the incompressibility of the rubber,
Radial stress is initially compressive, but becomes it has to find a new shape. It causes radial stress
tensile after 8 s of the simulation as the cupping reversal from a tensile stress to a compressive stress
begins to grow more rapidly in this region. The fric- in the aluminium workpiece. This reversal of radial
tion between the blank and the rubber pad causes the stress continues until the end of the process at t ¼ 34 s
initial compressive radial stress that helps to push in which enough compressive stress is applied to the
material into the die. This friction can be useful, if workpiece to cause it to take the die shape.

Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture JEM1453 Ó IMechE 2009

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

 Numerical simulation of sheet stamping process using flexible punch 839

Fig. 16 Strain distribution during the rubber-pad forming simulation with natural rubber

The distribution of strain in node A during the thinning is compared with experimental measure-
process is shown in Fig. 16. It can be seen from ments to validate the FE model. The main conclu-
the figure that radial strain is positive for the entire sions of this research are summarized below.
simulation, while axial strain is compressive. The
1. The rubber-pad forming process contains three
reason for the compressive axial strain is the friction
stages: flexible punch self-deformation; blank
condition between the blank and rubber pad, which
deformation to reach the bottom of the rigid die
causes material to be pushed into the die during the
and filling of the die cavity with blank until it fits
forming process. A combination of positive radial
the rigid die.
strain and negative axial strain causes thinning of
2. At full-load application, all the zones are under
specimen during the process.
tensile stress resulting in a thinning of the blank.
Even though the radial stress remains relatively
3. The maximum thinning in the specimen is
constant between t¼30 to 33 s as shown in Fig. 15, the
11.7 per cent using natural rubber and decreases
radial strain increases abruptly from 0.015 to 0.05 as
to 10.6 per cent using silicon rubber as the flex-
shown in Fig. 16. This indicates that the strain at node
ible punch. The punch velocity does not have
A (see Fig. 6(b) ) is now being influenced by the
significant effect on the thinning phenomena.
material flow occurring elsewhere, particularly at
4. The maximum punch load necessary to stamp
bend points 2 and 4 (see Fig. 6(a) ). The plastic flow at
the blank is 164.8 kN for natural rubber and 67.3
points 2 and 4 causes material to flow radially out-
kN for silicon rubber. This clearly shows that the
wards, which result in a thinning at point A. The
requirement for a high-capacity press can be
stress and strain analysis gives us better insight into
overcome by using silicon rubber. However, the
the deformation process that can be used for further
silicon rubber has a shorter lifetime compared to
analysis and optimization of the rubber-pad forming natural rubber and can not be used to form
process. specimens with sharp edges.
5. Shore A hardness of rubber does not have a sig-
nificant effect on the thickness thinning of the
4 CONCLUSIONS formed part.
6. As the rubber-pad thickness decreases, the thin-
In this paper the rubber-pad forming experiments ning of the formed blank increases. However this
were performed to stamp aluminium blanks. FE increase in thinning is not significant.
simulations were carried out in ABAQUS/Standard to 7. Higher friction between rubber and blank would
analyse the process. The FE prediction of specimen help in pulling more material into the die and

JEM1453 Ó IMechE 2009 Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013

840 M Ramezani, Z M Ripin, and R Ahmad

hence reduce thinning of the formed blank. 6 Dirikolu, M. H. and Akdemir, E. Computer-aided
Higher friction between blank and rigid die make modelling of flexible forming process. J. Mater. Process
the metal sheet more difficult to flow into the die, Technol., 2004, 148, 376–381.
resulting in more stretching and thinning of the 7 Ramezani, M., Ripin, Z. M., and Ahmad, R. Computer-
aided modelling of friction in rubber-pad forming pro-
blank.
cess. J. Mater. Process Technol., 2009. DOI: 10.1016/
8. The deformation mechanism that forms the
j.jmatprotec.2009.01.015.
blank is a combination of radial tension and axial 8 Peng, L., Hu, P., Lai, X., Mei, D., and Ni, J. Investigation
compression during the forming process. of micro/meso sheet soft punch stamping process –
simulation and experiments. Mater. Des., 2009, 30, 783–
790.
9 Ramezani, M., Ripin, Z. M., and Ahmad, R. Modelling
of kinetic friction in V-bending of ultra-high-strength
REFERENCES
steel sheets. Int. J. Adv. Mf. Technol., 2008. DOI: 10.1007/
s00170-008-1450-4.
1 Thiruvarudchelvan, S. Elastomers in metal forming: a 10 Ramezani, M., Ripin, Z. M., and Ahmad, R. A static
review. J. Mater. Process Technol., 1993, 39, 55–82. friction model for tube bulge forming using a solid
2 Browne, D. J. and Battikha, E. Optimisation of alumi- bulging medium. Int. J. Adv. Mf. Technol., 2008. DOI:
nium sheet forming using a flexible die. J. Mater. Process 10.1007/s00170-008-1708-x.
Technol., 1995, 55, 218–223. 11 Mooney, M. A theory for large elastic deformation.
3 Thiruvarudchelvan, S. The potential role of flexible J. Appl. Phys., 1940, 11, 582–597.
tools in metal forming. J. Mater. Process Technol., 2002, 12 MacDonald, B. J., and Hashmi, M. S. J. Three-
122, 293–300. dimensional finite element simulation of bulge form-
4 Thiruvarudchelvan, S. and Tan, M. J. Recent deve- ing using a solid bulging medium. Finite Element
lopments in friction-assisted sheet metal forming pro- Anal. Des., 2001, 37, 107–116.
cesses. J. Mater. Process Technol., 2005, 167, 161–166. 13 Girard, A. C., Grenier, Y. J., and MacDonald, B.
5 Sala, G. A numerical and experimental approach Numerical simulation of axisymmetric tube bulging
to optimise sheet stamping technologies: part II – using a urethane rod. J. Mater. Process Technol., 2006,
aluminium alloys rubber-forming. Mater. Des., 2001, 22, 172, 346–355.
299–315.

Proc. IMechE Vol. 223 Part B: J. Engineering Manufacture JEM1453 Ó IMechE 2009

Downloaded from pib.sagepub.com at Universitats-Landesbibliothek on December 4, 2013