Journal of Materials Processing Technology 82 (1998) 156 – 164

Hydraulic pressure enhancement of the deep-drawing process to

yield deeper cups
S. Thiruvarudchelvan *, H.B. Wang, G. Seet
School of Mechanical and Production Engineering, Nanyang Technological Uni6ersity, Nanyang A6enue, Singapore 639798, Singapore

Received 12 July 1997

Abstract

In a novel process recently developed by the present authors, high hydraulic pressure proportional to the punch force generated
in the tooling augments the drawing action of the punch, provides blank holding, and lubricates the blank – tool interfaces,
yielding deeper cups. This hydraulic-pressure-assisted deep-drawing technique, which enables large draw ratios, is introduced in
this paper. Hydraulic pressure proportional to the punch force provides the blank-holding force, which increases from zero at the
beginning to a maximum and then drops to zero at the end of drawing — somewhat similar to the critical minimum required to
suppress wrinkling. This helps to minimize the frictional resistance at the flange at the initial critical stage of drawing. The
hydraulic pressure is also applied on the periphery of the flange of the cup, so that the drawing is performed in a push–pull
manner, enabling higher draw ratios than those achieved in the conventional deep-drawing process. The process has inherent
advantages such as the automatic coordination of the punch force, the hydraulic pressure and the blank-holding force, and low
friction between the blank and the tooling, as a high-pressure liquid lubricates these interfaces. This process needs only a
single-action press unaided by any cushion, thereby reducing the capital investment needed. The principle of the process is
explained in this paper. Results of experiments conducted using annealed aluminum blanks of thickness 1 mm to draw cups of
36 mm diameter at draw ratios of up to 3.5 are reported here. A finite-element analysis of the process has also been carried out.
Measured hydraulic pressures are compared with those calculated using a simple analytical method and those from the
finite-element analysis. Strain and thickness measurements taken with cups drawn by this process are also presented. © 1998
Elsevier Science S.A. All rights reserved.

Keywords: Deep-drawing process; Hydraulic pressure enhancement; Punch force

1. Introduction being drawn can be kept low, thereby increasing the

draw ratio to a somewhat higher value.
The deep-drawing process, which is used widely in In order to overcome the limitations of the process,
industry, produces cups at draw ratios of less than and to obtain good quality cups, new variations have
about 2.2. Various techniques are adopted to achieve been introduced over the years, notable in this category
higher draw ratios. In conventional deep drawing, the being hydroforming [2], hydro-mechanical forming [3]
process parameters are selected carefully to enable the and counter-pressure deep drawing [4]. Each of these
drawing of good quality cups at the maximum possible processes has distinguishing features and limitations [5].
draw ratios. Improvements in the tooling, proper lubri- Various investigations have been conducted to improve
cation of the die and optimum design of the punch and the processes [6]. They are already employed in industry
die radii, may increase the limiting draw ratio slightly. in specialized applications. Hillier used a simple ap-
By controlling the blank-holding force at each stage of proach to study the mechanics of these processes [7].
drawing [1], frictional resistance at the flange of the cup El-Sebaie analyzed the mechanics of hydro-mechanical
deep drawing, incorporating the effects of work harden-
* Corresponding author. Tel.: + 65 7995506; fax: + 65 7911859; ing and anisotropic material properties [8]. Yang et al.
e-mail: [email protected] [9] carried out a finite-element analysis to determine the

0924-0136/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.

PII S0924-0136(98)00035-1
 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164 157

effects of fluid pressure on the deformation in hydro-

mechanical deep drawing.
In the above processes, there are some common
phenomena that are responsible for the increased limit-
ing draw ratios achieved. The hydraulic-pressure-aug-
mented deep-drawing process investigated in this paper,
whilst having some common features with the above
processes, has also distinctly new features, these fea-
tures including the following: (i) automatic generation
of the blank-holding force; (ii) simplicity of the tooling
and its operation, needing only a single action press.
Also, the present process does not require a rubber
diaphragm or a seal on the flange, or result in large
amounts of leakage of fluid in the tooling, thereby
making the process more attractive than the above
processes. The objective of present research is to refine
and develop the present process, and determine the
correlation between the forming parameters of the pro-
cess. Theoretical- and finite-element analyses have also
been performed to gain more understanding of the
mechanics of the process.

2. The principle of the process

An illustration of the basics of the process is shown

in Fig. 1. When the press ram descends to start the
drawing process, the fluid in the device is pressurized.
The top-end diameter of the punch subjected to the
pressure is smaller than the diameter of the punch

Fig. 2. Tooling for the hydraulic-pressure-augmented deep-drawing

process.

drawing the cup, and the pressure generated is high.

The punch and the blank holder are subjected to the
same hydraulic pressure. The periphery of the blank
also receives the same hydraulic pressure through the
longitudinal holes on the blank holder. Thus, whilst the
punch is pulling the cup into the die cavity, the hy-
draulic pressure is pushing the flange of the cup radially
inwards. The hydraulic pressure acts on a larger area
on the top surface of the blank holder than it does on
its bottom surface at every stage of drawing. Hence, the
hydraulic pressure applies a net downwards force on
the blank, this constituting the blank-holding force.
The high-pressure fluid applied at the periphery of the
flange lubricates the flange and the die radius. How-
ever, significant leakage of fluid is prevented as the
blank holder forms a ‘seal’ based on the unsupported-
area principle of the sealing of high-pressure fluid. If
the volume of the fluid in the system is large, then the
cup will emerge at high speed at the end of drawing,
Fig. 1. Schematic diagram of the hydraulic-pressure-augmented deep- when the energy in the high-pressure fluid is released:
drawing process. To reduce the speed, the fluid volume can be mini-
158 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164

mized, and also longitudinal holes fitted with valves can emerged at high speed, but after some measures taken
be fitted to cut off the pressurized fluid supply to the to minimize the volume of the fluid in the container,
flange periphery whenever the cup being formed accel- and by reducing the energy released at the end of
erates. A step is also introduced on the die at the top to drawing, the cups now emerge at reasonable speeds.
take up any excessive blank-holding force. Once the The hydraulic pressure increases as the deforming
flange of the cup thins by more than 5% (for example), load increases with the punch stroke. Simultaneously,
the blank holder touches this step and transfers any as the blank moves radially inwards, the upwards pres-
extra force to the step instead of to the blank, in this sure acting on the bottom of the blank holder acts over
way excessive blank-holder force being avoided. a larger annular area, thereby balancing part of the
In the actual process, the drawing die is placed on the force acting downwards on the top of the blank holder.
press table and the rest of the tooling assembly is Thus, the blank-holder force does not become exces-
attached to the press ram, as shown in Fig. 2. Then the sive. The draw force in the present process can be
sheet metal blank is placed on the die, and the upper derived [10] assuming the blank-holder pressure, pH, to


assembly of the punch, the blank holder, the container be constant as:
and the punch holder, is lowered down to the blank.
The space between the blank holder and the punch FD = pdt emp/2 1.1s2 ln
r0 2mpH
+ (r0 − r1)− p
n
holder is then filled with hydraulic fluid. Air is bled r1 t
from the top port, which is then sealed. The press ram pdt 2s1
+ (1)
is brought down, pressurizing the fluid, which in turn 2rD
exerts force on the punch stem and the blank holder.
where d is the mean diameter of the cup, t is the
The pressurized fluid also reaches the periphery of the
original thickness of the blank, s2 is the mean yield
blank through four longitudinal holes on the blank
stress of the flange area, r0 is the current blank radius,
holder. The hydraulic pressure is thus applied on the
r1 is the internal radius of the cup, m is the coefficient of
periphery of the flange of the cup being drawn whilst
friction, p is the hydraulic pressure generated, s1 is the
the punch exerts force as in conventional drawing.
mean yield stress of the metal at the die throat, and rD
When the cup is fully drawn it drops through the die
is the die radius.
cavity, and the upper part of the assembly is moved up
From this equation, it is clear that the draw force
with the press ram, the tooling then being ready for
decreases as the hydraulic pressure p increases, and that
another cycle. In the early experiments, the cups
cups can be drawn at greater draw ratios than in
conventional deep drawing. The diameter of the punch
stem is chosen to ensure that a sufficiently high hy-
draulic pressure is applied on the periphery of the blank
to increase the possible draw ratio. This also results in
a satisfactory blank-holding force being applied on the
blank—a force that is large enough to prevent substan-
tial leakage of hydraulic oil through the interface be-
tween the blank and tooling, and also one that
eliminates wrinkling. By decreasing the diameter of the
punch stem, the blank-holding force can also be in-
creased: during the drawing process it varies with the
hydraulic pressure. A free-body diagram with the forces
acting on the blank, the blank holder and the die, is
shown in Fig. 3. The fluid pressure, p, acts downwards
on the top surface of the blank holder. The pressure
acts upwards on the bottom surface of the blank
holder, where it is not in contact with the blank. Thus,
the net downwards force on the blank holder is:
pp(r 20 − r 24)
This force is balanced by the blank-holding force on the
blank:
pHp(r 20 − r 23)
Therefore:
Fig. 3. Free-body diagram showing forces on the blank, the blank
holder and the die. pHp(r 20 − r 23)= pp(r 20 − r 24) (2)
 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164 159

Fig. 4. FEM model and simulation of the deformed cup.

i.e.: ple) and it is higher than the blank-holding pressure

2 2
needed in conventional drawing. The only concern is
pH r −r
\1
0 4
= (3) that the if the blank-holding pressure is too high and if
2 2
p r −r 0 3 the lubrication is not effective, then the frictional drag
In Eq. (3), the ratio of pH/p is greater than unity since on the blank will be excessive. Since the hydraulic
in the design of the tooling r4 is chosen to be smaller pressure is high, the lubrication should be efficient, as
than r3 for the process to operate at a reasonably high in hydrostatic extrusion. To reduce the possibility of
value of pressure. Also, since pH is greater than p, there excessive blank-holding force, a step of height 0.95 mm
will be no significant leakage (unsupported-area princi- is incorporated at the periphery of the die. If the
blank-holding force becomes excessive, the blank will
be squeezed from 1 mm to a thickness of 0.95 mm,
thereafter the excess force being transferred to the die
rather than to the flange of the cup. The effect of the
step is neglected in the above analysis.

3. FEM simulation of the process

The present process is pressure-controlled, and it has

to be formulated differently from the usual displace-
ment-controlled formulation for deep drawing. The
process has been subjected to an elastic–plastic finite-
element analysis using the commercial FEM code
MARC, which is a general-purpose finite-element code
Fig. 5. Photograph of the tooling assembled on an Enerpac press. for static and dynamic analysis. Implicit time integra-
160 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164

4. Experimental work

Experimental work was conducted using the tooling

(Fig. 2) assembled on an Enerpac 200 ton press, shown
in Fig. 5. Commercially pure aluminum blanks
(Al1100-H14) of thickness 1 mm and of diameters
80–126 mm were tested. The blanks were annealed at a
temperature of 350°C for 2 h and allowed to cool in the
furnace. After annealing, the blanks were emery-pa-
pered to a smooth finish to remove oxide scale and
scratches before drawing.
A punch and blank holder set was manufactured
with the punch stem having a diameter of 12 mm,
which would amplify the pressure by a factor of 9, as
the punch diameter is 36 mm. All experimental work
and theoretical analysis reported in this paper are based
on this design of the punch. To solve the excessive
thinning of the flange of the cup being drawn during
the initial trials, caused by excessive blank-holding
force being applied, the top face of the die was ma-
chined to a depth of 0.95 mm over a diameter of 126
mm. Thus, the outer annular area of the die face
(between 126 and 132 mm diameter) will take up any
extra blank-holding force once the flange thickness
Fig. 6. Hydraulic pressure generated for various draw ratios (experi-
reduces to 0.95 mm. Details of the early work on the
mental).
development of the tooling are reported elsewhere [12].
tion was used in the analysis. The MARC updated
Lagrangian procedure was utilized together with elasto-
plastic material, large displacement and isotropic hard-
ening options. The analysis of the deep-drawing process
was based on axisymmetric conditions. Only the right-
half of the tooling and the blank were modeled. Fig. 4
shows the deformed shape of the finite-element mesh of
the model at an intermediate stage of drawing.
Owing to the blank-holding force being dependent on
the hydraulic pressure, the blank holder was treated as
a deformable object with very high initial yield
strength. Thus, distributed pressure on the blank holder
can be simulated in the model. For the same reason, the
punch was also treated as a deformable object. The die
was modeled as a rigid surface. The contact option in
MARC was used to deal with the contact problems
between the blank and the tooling. The four-node
axisymmetric quadrilateral continuum element, element
10, was selected. The true-stress – strain curve of the
material of the blank was taken as [11]:

s̄ = 130(ō + 3.8× 10 − 4)0.215 MN m − 2 (4)

The stress–strain behavior was defined by a user sub-

routine. The von Mises yield criterion was adopted.
Rate dependence was not considered. The coefficient of
friction between the cup and the punch was taken to be
0.2. The friction on the flange and that at the die radius
were assumed to be negligible in the Coulomb friction Fig. 7. Hydraulic pressure generated for various draw ratios (analyti-
model. cal).
 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164 161

The part of the pressure profile up to the maximum

pressure in Figs. 6 and 7 are quite similar to each other.
Ref. [12] gives the maximum hydraulic pressures mea-
sured during the drawing of the cups. The blank-hold-
ing force is calculated using the measured fluid
pressure. The blank-holding force thus calculated is an
over-estimate, as part of the force is transferred to the
die surface rather than to the blank. The maximum
hydraulic pressures for different blank radii calculated
using Eq. (1) are plotted together with the experimental
values in Fig. 8. It can be observed that the theoretical
and the experimental values are quite close to each
other. The pressure generated when the flange radius is
reduced from the initial value of 50 mm, whilst drawing
a blank of diameter 100 mm to a cup of diameter 36
mm, from the simulation of the present process using
FEM is shown in Fig. 9. The maximum hydraulic
pressure of 77.0 MPa obtained from the FEM simula-
tion compares with the measured value of 75 MPa. The
hydraulic pressure calculated using the FEM is plotted
together with the analytical value calculated using Eq.
(1). From Fig. 9 it can been seen that the two results
are quite close except at small deformation when the

Fig. 8. Maximum hydraulic pressure generated for different blank

radii.

The hydraulic pressure generated in the tooling was

monitored continuously with the punch stroke. After
the pressure reaches the maximum value, the cup starts
to accelerate to a high speed and the measurement of
the pressure and the stroke is difficult, as the transduc-
ers used are not capable of responding to dynamic
situations. Fig. 6 shows the measured pressure versus
the punch stroke for different draw ratios. It may be
noted that the dashed lines are extrapolations of the
pressure profile when the cup accelerates, the transduc-
ers being unable to respond. However, in cases where
the cup was not completely drawn, but was left with a
residual flange, the pressure and the stroke were
recorded fully. The problem of the cup accelerating was
also not present in the interrupted tests, where the
drawing was stopped at various values of the punch
stroke. However, in these tests the pressure did not
reach the maximum value. The strains on the cup were
measured from the distortions of the grids of circles
etched on the blank before drawing.

5. Results and discussion

Fig. 7 shows the pressure profile with stroke, calcu- Fig. 9. Analytical- and FEM-determined values of the pressure
lated using Eq. (1) for various values of the draw ratio. generated during the drawing of a cup at a draw ratio of 2.8.
162 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164

Fig. 10. The strain distribution calculated by the FEM simulation of the process for a draw ratio 2.8.

FEM gives larger values for the pressure. This may be in the flange associated with the conventional deep-
because elastic strains are considered in the FEM simu- drawing process is modified here due to the presence of
lation, whilst they are not included in Eq. (1). a large blank-holding force. The flange has uniform
The strains on the cup calculated from the FEM thickness because of the presence of the recess on the
simulation with a blank diameter of 100 mm are shown die surface that prevents the blank holder from going
in Fig. 10. It is found that the portion at the bottom of down further after the flange deforms by about 0.05
the cup has undergone some thinning, whilst the rest of mm. Fig. 11 shows the strains on the cup wall measured
the cup has undergone some thickening. Also, there is using the grid etched on the blank (draw ratio 3.5), the
slight thinning near the brim of the cup. The thickening strain variation being similar to that in simple deep

Fig. 11. Measured true strain distributions along the center line for a draw ratio of 3.5.
 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164 163

Fig. 12. Measured thickness strain distributions of the drawn cup for different draw ratios.

drawing, except that the strains involved in the present The maximum hydraulic pressures measured during the
process are greater. Fig. 12 shows the cup wall thick- tests agree well with those calculated using a simple
ness measured when drawing cups at the three draw equation. The hydraulic pressures at different values of
ratios 3.0, 3.3 and 3.5. It can be seen that the wall the punch stroke calculated by the FEM simulation are
thickness is quite uniform except for the area near the also found to agree well with those measured during the
punch nose radius. The thinning is maximum at this drawing experiments. The strains calculated by the
point and increases as the draw ratio increases. Fig. 13 FEM simulation show some degree of thinning at the
is the photograph of the cup drawn with a draw ratio bottom of the cup and also slight thinning near the
of 3.5. brim of the cup. The measured strains also show a
similar trend. The measured thickness variations of the
cup wall are small, except for the area near the punch
6. Conclusions nose, where the maximum thinning occurs.

The present hydraulic-pressure-augmented deep-

drawing process has proven to be capable of drawing References
cups successfully from aluminum blanks of thickness 1
mm at draw ratios of up to 3.5 on a single-action press. [1] S. Thiruvarudchelvan, W.G. Lewis, Deep drawing with blank
holder force approximately proportional to the punch force,
Trans. ASME J. Eng. Ind. 85 (1990) 278 – 285.
[2] S. Yossifon, J. Tirosh, On the permissible fluid-pressure path in
hydroforming deep drawing processes — Analysis of failures and
experiments, Trans. ASME J. Eng. Ind. 110 (1988) 146–152.
[3] B. Larsen, Hydromechanical forming of sheet metal, Sheet Met.
Ind. (1977) 162 – 166.
[4] H. Amino, K. Nakamura, T. Nakagawa, Counter-pressure deep
drawing and its application in the forming of automobile parts,
J. Mater. Process. Technol. 23 (1990) 243 – 265.
[5] S. Thiruvarudchelvan, Hydraulic pressure-assisted cup drawing
processes — a review, ASME Yearbook, Singapore Chapter,
1992, pp. 17 – 24.
[6] K. Nakamura, T. Nakagawa, Fracture mechanism and fracture
control in deep drawing with hydraulic counter pressure. Studies
on hydraulic counter pressure forming, J. Jpn. Soc. Technol.
Plast. 25/26: (284) 831 – 838; (298) 1110 – 1116; (290) 284–290;
(299) 1188 – 1194.
[7] M.J. Hillier, The mechanics of some new processes of cup
drawing, J. Appl. Mech. ASME 36 (1969) 304 – 309.
[8] M.G. El-Sebaie, Plastic instability conditions when deep-drawing
into a high pressure medium, Int. J. Mech. Sci. 15 (1973)
485 – 501.
[9] D.Y. Yang, J.B. Kim, D.W. Lee, Investigation into manufactur-
ing of very long cups by hydromechanical deep drawing and
ironing with controlled radial pressure, Ann. CIRP 44 (1995)
Fig. 13. Photograph of the cup drawn at a draw ratio of 3.5. 255 – 258.
164 S. Thiru6arudchel6an et al. / Journal of Materials Processing Technology 82 (1998) 156–164

[10] K. Lange, Handbook of Metal Forming, McGraw–Hill, New cess. Technol. 33 (1992) 175 – 194.
York, 1985, pp. 20.1–20.69. [12] S. Thiruvarudchelvan, A novel hydraulic-pressure augmented
[11] S. Yossifon, K. Sweeney, T. Altan, On the acceptable blank- deep-drawing process for high draw ratios, J. Mater. Process.
holder force range in the deep drawing process, J. Mater. Pro- Technol. 54 (1995) 355 – 361.