Int J Advanced Design and Manufacturing Technology, Vol. 6/ No.

1/ March - 2013 87

Study on the Effect of Forming

Parameters in Sheet
Hydrodynamic Deep Drawing
Using FEM-based Taguchi
Method

D
B. Zareh*
Department of Mechanical Engineering,
Babol Noshirvani University of Technology, Iran
E-mail: [email protected]
*Corresponding author

S
A. H. Gorji & M. Bakhshi & S. Nourouzi
I
f
Department of Mechanical Engineering,
Babol Noshirvani University of Technology, Iran

o
Received: 9 Mars 2012, Revised: 14 June 2012, Accepted: 29 July 2012

e
Abstract: In this paper, a FEM-based Taguchi method is used to determine the
effects of forming parameters on the quality of part formability in the process of
hydrodynamic deep drawing assisted by radial pressure. Four important forming

v
parameters, fluid pressure, friction coefficient at blank/punch interface, die

i
entrance radius and the amount of gap (g) between die rim block and blank holder
are considered in this investigation. In order to have more comprehensive study,
three different workpieces are used as the case studies. Three-dimensional FE

c hmodels are developed for simulating the forming process. After experimental
validation, these models are used for performing the set of experiments designed
by Taguchi’s L_9 orthogonal array. The signal-to-noise (S/N) ratio and the analysis
of variance (ANOVA) techniques are used to calculate the contributions of each of

A r the mentioned parameters to the output characteristic. The results indicate that
fluid pressure and friction coefficient are the most influential parameters. Also, die
entrance radius and the amount of gap (g) have considerable effect on the part
formability. The obtained results may provide useful guidance on determining
forming parameters.

Keywords: Analysis of Variance, Finite Element Simulation, Sheet

Hydroforming, Taguchi Method
Reference: Zareh, B., Gorji, A, H., Bakhshi, M. and Nourouzi, S., “Study on the
Effect of Forming Parameters in Sheet Hydrodynamic Deep Drawing Using FEM-
based Taguchi Method”, Int J of Advanced Design and Manufacturing
Technology, Vol. 6/ No. 1, pp.87–99.

Biographical notes: B. Zareh is PhD student of Mechanical Engineering at

University of Tabriz, Iran. His current research interests include Advanced Metal
Forming Processes and Dynamic Behavior of Materials. A. Gorji is Assistant
Professor of Mechanical Engineering at Babol Noshirvani University of
Technology, Iran. M. Bakhshi is full Professor of Mechanical Engineering at
Babol Noshirvani University of Technology, Iran. He received his PhD in
Mechanical Engineering from Birmingham University, UK. S. Nourouzi is
Assistant Professor of Mechanical Engineering at Babol Noshirvani University of
Technology, Iran. He received his PhD in Surface Engineering from University of
Limoges, France.

www.SID.ir
© 2013 IAU, Majlesi Branch
88 Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March– 2013

1 INTRODUCTION process parameters on hydromechanical deep drawing

process and gave the conclusion that increasing the
Sheet hydroforming is a recently developed technology friction between blank and die or blank and blank-
that uses a pressurized fluid medium to deform a holder decreases the LDR value. Singh and Kumar [17]
workpiece. This process has gained an increasing investigated the effect of pre-bulging pressure and cut-
interest during the last couple of years, especially as off pressure on thickness distribution and surface finish
application in the manufacturing of some components in hydro-mechanical deep drawing. However, the
for automotive, aerospace, and electrical appliances [1], pervious researchers have considered the effect of
[2]. Sheet hydroforming technology offers several forming parameters independently and consideration of
technological advantages in comparison to the influence of the forming parameters simultaneously
conventional stamping processes, such as a higher in sheet hydroforming process has not been found in
drawing ratio, better surface quality, less springback, the literature. Also, the previous works have studied the
fewer secondary operation, better dimensional accuracy effect of forming parameters only on the
and the capability of forming complicated-shaped sheet hydroformability quality of a workpiece with a specific
metal parts [3], [4]. Although this process has many
advantages and good application prospects, the usage
of liquid in the die cavity makes this process more
complicated than conventional deep drawing. Also, in
geometry.

I D
The Taguchi method has been widely applied in
various industrial fields including the metal forming
area in order to design a robust manufacturing process
comparison to the conventional deep drawing, further
parameters affect the formability of sheet metal during
the sheet hydroforming process. For wide application
of this technology, there is a need for a fundamental
S
and determine optimal design parameters [18]–[20].
The Taguchi method adopts a set of standard
orthogonal

f arrays to determine parameters

configuration and analysis results. These kinds of
understanding of the influence of forming parameters
and their degree of importance on sheet
hydroformability, which will be very helpful for
spreading this technology in industry and for adding
more knowledge into its database.
e o
arrays use a small number of experimental runs but
obtain maximum information and have high
reproducibility and reliability. In this paper, finite
element method and Taguchi method are combined to
analyse the effects of forming parameters on the quality
In recent years, several new sheet hydroforming

i v
processes have been introduced such as aquadraw deep
drawing, hydromechanical deep drawing,
of part formability and determine the optimal
combination of forming parameters in sheet
hydroforming process. The process of hydrodynamic

h
hydrodynamic deep drawing (HDD), hydraulic deep deep drawing assisted by radial pressure (HDDRP) is
drawing with counter pressure, hydro-form, sheet employed as an example to apply the proposed

r c
hydroforming with a movable die, hydrodynamic deep
drawing assisted by radial pressure (HDDRP) and
hydromechanical deep drawing with uniform pressure
on the blank [4-10]. Also, the effect of forming
approach. In order to have more comprehensive study,
three different workpieces are considered as case
studies in this investigation. This type of analysis is not
available in the literature and will be very helpful to the

A
parameters on the sheet hydroforming process has been development and practical application of this
studied by some researchers using experiments, technology for a wide range of sheet metal
analytical models and finite element simulations. Hsu components.
and Hsieh [11] developed an analytical model to
predict the upper bound and lower bound of the
permissible fluid pressure in hemispherical-cup 2 METHODOLOGY
hydroforming process. They found out that the material
properties which include the strain hardening and Taguchi techniques have been utilized widely in
normal anisotropic parameter tend to shift the working engineering analysis to optimize the performance
zone. Zhang et al. [12] showed that material anisotropy characteristics within the combination of design
and pre-bulging pressure have a remarkable influence parameters. The Taguchi method considers three stages
on the final product quality. Lang et al. [13-15] carried in a process development: (1) system design, (2)
out a series of research works on the effects of the key parameter design, and (3) tolerance design [21]. The
process parameters such as pre-bulging, punch surface focus of the system design is on determining the
roughness and fluid pressure during hydro-mechanical suitable working levels of design factors. Parameter
deep drawing with uniform pressure onto the blank. design seeks to determine the factor levels that produce
Fazli and Dariani [16] used FEM to study the effect of the best performance of the product/process under

© 2013 IAU, Majlesi Branch www.SID.ir

Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March - 2013 89

study. The optimal condition is selected in a way that

the influence of uncontrollable factors causes minimum
variation of system performance. Tolerance design is a
way to fine-tune the results of the parameter design by
tightening the tolerance of factors with significant
influence on the product. Among these stages,
parameter design is the key step in the Taguchi method
to achieve high quality without increasing the costs. To
obtain high forming performance in the sheet
hydroforming process, the parameter design approach
proposed by the Taguchi method is adopted in this
paper.
The purpose of this study is to investigate the influence
of forming parameters on the hydroformability to
improve the hydroformed sheet quality. In order to
achieve above target, first, the objective function is
defined and the forming parameters are selected, and
the appropriate orthogonal array is constructed. Then,
I D
the three-dimensional finite element (FE) model is
developed for simulating the HDDRP process using
dynamic explicit, commercial code, Abaqus 6.7.1. In
order to validate the FE models, thickness distribution
Fig. 1

f S
Schematic representation of hydrodynamic deep
drawing assisted by radial pressure.
curves and punch load-stroke curves of the
experimentally formed cups are compared with the
predicted values. These validated models are then used
to carry out virtual experiments for each set of forming
parameters designed by Taguchi’s standard orthogonal

e o
v
4 SELECTION OF THE FORMING PARAMETERS
array. Then, the virtual experimental results are AND CONSTRUCTION OF THE ORTHOGONAL

h i
transformed into the Taguchi’s signal-to-noise (S/N)
ratios and the optimal parameters are obtained. Finally,
Statistical analysis of variance (ANOVA) is performed
to see which parameters are significant.
ARRAY

Sheet hydroforming process and the quality of the

formed parts can be influenced by many forming

c
parameters. Generally, there are three categories of
parameters influencing hydroformability, i.e., geometry

A r
3 HYDRODYNAMIC DEEP DRAWING ASSISTED
BY RADIAL PRESSURE (PROCESS DESCRIPTION)

The process of hydrodynamic deep drawing assisted by

radial pressure (HDDRP) is shown in Fig. 1, which is
developed from the conventional hydrodynamic deep
parameters, material parameters, and process
parameters. Based on the previous study reported by
Sharma and Rout [23], for the entire range of a specific
material, the influence of material properties such as
strain hardening exponent and plastic anisotropy ratio
is less than other forming parameters. Moreover, in
drawing (HDD) through a little modification of the tool most cases for a specific workpiece, there is a little
setup. In the HDDRP process, when the punch goes flexibility in material selection. Hence, the process
down into the die cavity, the pressurized liquid in the parameters and geometry parameters are taken into
die cavity will push the blank tightly onto the punch account in this study. Among all process parameters
surface. In the meantime, the liquid in the die cavity and geometry parameters, fluid pressure in the die
may flee out form the gap between the blank and the cavity (Ps), interfacial friction condition between blank
die. At the same time, because the gap g where the and punch surface and die entrance radius are the most
liquid flees out is very small, a liquid pressure exists important parameters that can affect the sheet
around the blank rim. This is different from hydroformability during the sheet hydroforming
conventional HDD process [22]. The HDDRP process process. Also, in HDDRP process the amount of gap
has many forming parameters involved in affecting its (g) has a great influence on the process window.
hydroformability. So, in the present research, this Therefore, these parameters are taken into account in
process is taken as an example to implement previously this investigation. To evaluate, three levels are selected
mentioned procedure. for each of the above mentioned parameters. Table 1

www.SID.ir
© 2013 IAU, Majlesi Branch
90 Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March – 2013

shows the selected parameters and their levels used in 5 FINITE ELEMENT SIMULATION MODEL
FE simulation. For four factors with three levels for
each, the experimental layout of L_9 orthogonal array, In this study, finite element simulation is used as a
which has 9 rows corresponding to the number of tests substitute tool for experimental tests. The commercial
(8 degree of freedom), is selected for present research finite element software, ABAQUS/Explicit 6.7 is used
according to Taguchi’s suggestion [24]. Table 2 shows for simulation of the HDDRP process. The FE models
the L_9 orthogonal array in which 9 runs are carried are shown in Fig. 2 in an exploded view. As shown in
out to investigate the effects of the four selected the Fig. 2, only a quarter of the blank and the tool
factors. components are modeled due to symmetry. The
dimension of the initial blank and the tool components
Table 1 Parameters and their levels are listed in Table 3. According to Fig. 1, in HDDRP
Parameters Level 1 Level 2 Level 3 process a radial pressure is loaded onto the blank rim to
A: Fluid pressure (MPa) 10 20 30 push the blank forward. But in general finite element
softwares, the pressure vector vertical to the normal of

D
B: Friction coefficient (µ) 0.08 0.14 0.20
C: Gap (g) (mm) 0.00 0.05 0.10 one shell element cannot be supplied to the element.

I
D: Die entrance radius (mm) 3.00 4.00 5.00 So, in order to apply the radial pressure onto the blank
rim, the blank is modeled as a solid deformable body
and is meshed with eight-node solid elements C3D8R.

Ex. no.
1
2
Table 2 Taguchi’s L9 orthogonal array
A
1
1
B
1
2
C
1
2
D
1
2

f S
The applied material is St14 steel sheet with 1 mm
thickness. The necessary blank material properties
which are obtained from tensile test are given in Table
4. The tool components are modeled as rigid surfaces

o
3 1 3 3 3 using discrete rigid surface elements R3D4. The
4 2 1 2 3 Coulomb’s friction law model is applied to define the
5 2 2 3 1 friction contact condition between the interfaces. The
6
7
8
2
3
3
3
1
2
1
3
1

v
2
2
3
e
friction coefficient at blank/blank holder and blank/die
interfaces is considered to be 0.05. According to the
input parameters sets designed by Taguchi’s L_9
9 3 3 2

h i
Considering the widespread applications of cylindrical,
1 orthogonal array, the friction coefficient at blank/punch
interface is taken as µ=0.08, 0.14 or 0.20.

Table 3 Dimension of the tool components

r c
hemispherical and parabolic work pieces in the fields of
aviation, aerospace, automobile industry, etc., in this
work, hydroforming of these parts are considered as the
case studies to investigate the effects of mentioned
Parameter
Punch diameter
Die inside diameter
Die entrance radius
Value (mm)
41.8
44
3

A
parameters. Hence, the set of simulations designed by
Taguchi’s L_9 orthogonal array are carried out for each
of the three considered case studies. Therefore, 27
finite element simulation runs are performed in this
study.
Although there are many different proposed criteria for
Blank holder inside diameter
Blank holder entrance radius
Initial blank diameter (mm)
41.9
3
80

Table 4 Mechanical properties of the St14 steel sheet

predicting fracture in sheet metal formed parts, there is Parameter Value
no standard approach. Therefore, the commonly used Young modulus, E (GPa) 210
thinning ratio criteria is used here as a measure of Yield stress, σ (MPa) 175
Strain-hardening exponent, n 0.36
forming quality. The thinning ratio is defined by:
Strength coefficient, K (MPa) 648.53
Poisson’s ratio υ 0.30
t −t
Thinning ratio (%) = 0 1 × 100 (1) Normal anisotropy R 1.53
t
0
The amount of gap (g) is one of the parameters that is
where t0 is the original thickness of the sheet and t1 is investigated in this study. This parameter directly
the critical thickness of the hydroformed cup. affects the amount of radial pressure loaded onto the

© 2013 IAU, Majlesi Branch www.SID.ir

Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March - 2013 91

I D
Fig. 2

f S
The exploded view of the FEM models (one-quarter geometry).

blank rim. Based on reference [22] for the specified

range of gap (g) following Eq. (2) is used to calculate
this pressure in the simulations:

e o
In HDDRP process, liquid pressure in the die cavity
varies during the forming process. Also, the liquid
pressure distribution under flange region is non-
uniform, shown in Fig. 3. So, in the FE model the

P =
r
6η Q ⎛ R ⎞
ln ⎜ ⎟
πg 3 ⎝ a ⎠

i v (2)
pressure distribution under the blank must be defined as
a function of nodes coordinates and time.
ABAQUS/Explicit can be applied in combination with
the VDLOAD subroutine. This subroutine can be used

h
Where Pr is the liquid pressure along the rim of the

c
blank, η the dynamic viscosity of the liquid, Q the flow
rate, g the gap, R the external radius for the gap and a is
to define the variation of the distributed load magnitude
as a function of position, time, velocity, etc [25]. Thus,
VDLOAD subroutine is used for modeling the fluid

r
the internal radius for the gap. These parameters are
shown in Fig. 3. The appropriate levels for the gap g
are selected in a way that the amount of radial pressure
varies from Pr؄Ps to Pr؄0.

A
pressure.

6 EXPERIMENTAL PROCEDURE AND

VALIDATION OF THE FE MODEL

A Denison Mayes Group universal testing machine

with 600-kN capacity is used in the experiments. Fig. 4
shows the experimental equipment and the
manufactured die set mounted on the test machine. The
pressure generating system is a hydraulic unit with a
maximum capacity of 35 MPa. The working pressure is
regulated by a pressure relief valve. Fig. 5 shows the
formed parts. In order to verify the validity of
developed FE model, the thickness distribution and
punch force variations of the formed parts in the
experimental tests are compared with those obtained
from the finite element simulations. These comparisons
Fig. 3 The fluid pressure distribution on the flange area are shown in Fig. 6 and 7, respectively. As can be
during the forming process [22]

www.SID.ir
© 2013 IAU, Majlesi Branch
92 Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March – 2013

Fig. 4 Experimental equipment and the manufactured die

D
set Fig. 5 Photograph of the hydroformed work pieces
obtained from experiments (a) parabolic cup, (b)
inferred from the figures, the finite element results are
in good agreement with the experimental results and
hence the developed FE model can be used as a proper
substitute tool for performing the set of experiments
designed by Taguchi’s L_9 orthogonal array.
S I
hemispherical cup, (c) cylindrical cup

o f
RESULTS AND DISCUSSION

Three different sheet metal parts, cylindrical cup, hemi-

spherical cup and parabolic cup were considered as the
case studies in this investigation. Finite element

v e
simulations were performed for each of the above
mentioned parts according to the arrangements of
Taguchi’s L_9 orthogonal array (Table 2). The results

h i obtained from finite element simulations were then

transformed into signal-to-noise (S/N) ratio. The
significance of the design factors were estimated by the
analysis of variance (ANOVA) statistical method.

r c 7.1. S/N Analysis

Taguchi’s method uses a statistical measure of
performance called signal-to-noise ratio (S/N), which is

A
logarithmic function of desired output to serve as
objective functions for optimization [24]. It is defined
as the ratio of the mean (signal) to the standard
deviation (noise). The ratio depends on the quality
characteristics of the product/process to be optimized.
There are three categories of performance
characteristics in the analysis of the S/N ratio: the
lower-the-better (LB), the higher-the-better (HB), and
the nominal-the-better (NB). The S/N ratio is expressed
as:

S / N = −10 log( MSD ) (3)

where MSD is the mean squared deviation from the

target value of the quality characteristic. For the case
of minimization of maximum thinning ratio, LB
characteristic needs to be used. The MSD for the lower-
the-better quality characteristic can be expressed as:

© 2013 IAU, Majlesi Branch www.SID.ir

Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March - 2013 93

cup.
1 n 2 (4)
MSD = ∑ y
n i =1 i

where yi is the measured value of the lower-the-better

quality characteristic and n is the total number of tests
for a trial condition.

I D
f S
e o
i v
c h
A r
Fig. 7 Comparison of the punch load-stroke curves
obtained by the experiment and FE simulation (a) parabolic
cup, (b) hemispherical cup, (c) cylindrical cup

Table 5 shows the finite element simulation results for

the maximum thinning ratio and its corresponding S/N
ratio for the three different hydroformed parts in 9 trial
conditions. Since the experimental design is
Fig. 6 Comparison of the thickness distribution curves orthogonal, the effect of each control factor on the S/N
obtained by the experiment and FE simulation (a) parabolic ratio at different levels can be separated out. The
cup, (b) hemispherical cup, (c) cylindrical
average S/N ratio of each parameter at levels 1 to 3, for

www.SID.ir
© 2013 IAU, Majlesi Branch
94 Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March – 2013

parabolic cup, hemi-spherical cup and cylindrical cup the process parameters is the level with the greatest S/N
are listed in Tables 6, 7 and 8, respectively. The ratio. According to the Fig. 8, the optimal process
corresponding main effect plots are also shown in Fig. parameters combination for minimization of the
8. Regardless of the category of the quality maximum thickness reduction of parabolic cup,
characteristic, a greater S/N ratio corresponds to better hemispherical cup and cylindrical cup are obtained as
quality characteristics. Therefore, the optimal level of A2B3C1D3, A2B3C1D3 and A1B3C1D3, respectively.

Table 5 Experimental results for maximum thinning and S/N ratios

Run.N Parabolic cup Hemispherical cup Cylindrical cup
Thinning ratio S/N ratio Thinning ratio S/N ratio Thinning ratio S/N ratio
1 0.28297 10.9651 0.13901 17.1390 0.12633 17.9696
2 0.27194 11.3102 0.13020 17.7077 0.11633 18.6857
3 0.25900 11.7338 0.12513 18.0527 0.11053 19.1303
4 0.26240 11.6205 0.11781 18.5763 0.11811 18.5536
5
6
7
8
0.24844
0.21898
0.27094
0.23736
12.0953
13.1917
11.3425
12.4917
0.10794
0.08486
0.13001
0.08993
19.3363
21.4259
17.7204
20.9212
I D
0.12728
0.11053
0.13769
0.11886
17.9044
19.1298
17.2218
18.4987
9

Parameter
0.23750 12.4867 0.09939

Level 1
20.0531

Level 2

f S 0.13411

Table 6 S/N response table for maximum thinning ratio (Parabolic cup)
Level 3
17.4503

|Max-Min|
A

B
Mean Th. ratio
Mean S/N ratio
Mean Th. ratio
Mean S/N ratio
Mean Th. ratio
0.27130
11.3364
0.27210
11.3094
0.24643

e o 0.24327
12.3025
0.25258
11.9658
0.25728
0.24860
12.1070
0.23849
12.4707
0.25946
0.02803
0.96615
0.03361
1.16138
0.01302
C

D
Mean S/N ratio
Mean Th. ratio
Mean S/N ratio
i v12.2162
0.25630
11.8490
11.8058
0.25395
11.9481
11.7239
0.25292
11.9487
0.49231
0.00338
0.09965

Parameter

c h
Table 7 S/N response table for maximum thinning ratio (Hemispherical cup)
Level 1 Level 2 Level 3 |Max-Min|
A

C
A r
Mean Th. ratio
Mean S/N ratio
Mean Th. ratio
Mean S/N ratio
Mean Th. ratio
Mean S/N ratio
0.13144
17.6332
0.12894
17.8119
0.10460
19.8287
0.10353
19.7795
0.10935
19.3217
0.11580
18.7790
0.10644
19.5649
0.10312
19.8439
0.12102
18.3698
0.02791
2.14633
0.02581
2.03198
0.01642
1.45888
Mean Th. ratio 0.11544 0.11503 0.11095 0.00448
D
Mean S/N ratio 18.8428 18.9513 19.1834 0.34059

Table 8 S/N response table for maximum thinning ratio (Cylindrical cup)
Parameter Level 1 Level 2 Level 3 |Max-Min|
Mean Th. ratio 0.11773 0.11864 0.13022 0.01249
A
Mean S/N ratio 18.5952 18.5293 17.7236 0.87157
Mean Th. ratio 0.12738 0.12082 0.11839 0.00898
B
Mean S/N ratio 17.9150 18.3629 18.5701 0.65512
Mean Th. ratio 0.11857 0.12285 0.12516 0.00659
C
Mean S/N ratio 18.5327 18.2299 18.0855 0.44722
Mean Th. ratio 0.12924 0.12152 0.11583 0.01340
D
Mean S/N ratio 17.7748 18.3457 18.7275 0.95275

© 2013 IAU, Majlesi Branch www.SID.ir

Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March - 2013 95

sum of the squared deviations SST from the total mean

S/N ratio can be calculated as [26]:

2 1⎡ m S 2
m ⎛S ⎞ ⎛ ⎞ ⎤ (5)
SS = ∑ ⎜ ⎟ − ⎢ ∑ ⎜ ⎟ ⎥
T
i = 1⎝ N ⎠ i m ⎢⎣i = 1 ⎝ N ⎠ i ⎦⎥

where m is the number of experiments in the

orthogonal array and (S/N)i is the S/N ratio of the ith
experiment. The sum of the squares due to the variation
from the total mean S/N ratio for the pth parameter is
expressed as:

D
⎛⎛ S ⎞ ⎞
⎜⎜ ⎟ ⎟
2
l ⎜⎝ ⎝ N ⎠ j ⎟⎠
SS = ∑
P
j =1

S
t

I
−
1 ⎡ m ⎛S ⎞ ⎤
⎢ ∑ ⎜ ⎟ ⎥
m ⎢⎣i = 1 ⎝ N ⎠ i ⎥⎦

where l is the number of the parameter levels (l = 3 in

(6)

o f
this study), j is the level number of this specific
parameter p, (S/N)j is the sum of the S/N ratio
involving this parameter p and level j, and t is the
repetition of each level of parameter p. The percentage
contribution of the pth parameter can be calculated as:

v e
P (%) =
P
SS
P × 100 (7)

i
SS
T

h
The results of ANOVA for the parabolic cup,
hemispherical cup and cylindrical cup are shown in

c
Tables 9 to 11, respectively. Based on the ANOVA
results given in Table 9, the friction coefficient at

A r blank/punch interface (50.39%) is the most significant

parameter influencing the maximum thinning ratio of
the parabolic cup followed by fluid pressure and the
amount of gap g. Die entrance radius has the least
effect on the maximum thinning ratio and is assumed to
be negligible and is pooled to form the error variance
estimate. In the case of hemispherical cup the fluid
Fig. 8 Main effect plots (a) parabolic cup, (b)
pressure (44.96 %) is found to be the most significant
hemispherical cup, (c) cylindrical cup
parameter influencing the maximum thinning ratio, as
shown in Table 10. Following this, the descending
7.2. Analysis of variance (ANOVA)
order of contributions are the friction coefficient at
The Taguchi experimental analysis provides the blank/punch interface and the amount of gap g. Die
information not only for the selection of an optimal entrance radius again has the least contribution and is
condition, but also for the evaluation of the relative pooled. Table 11 shows the results of ANOVA for the
importance of each factor for further studies. The cylindrical cup. It is seen that the fluid pressure (37.40
analysis of variance (ANOVA) procedure is used to %) and die entrance radius (36.51 %) have got the most
quantify the influence of the forming parameters on the significant influence on the maximum thinning ratio of
quality characteristic. ANOVA separates the overall the cylindrical cup while the friction coefficient and the
variation from the average S/N ratio into contribution amount of gap g have also got strong influence on the
by each of the parameters and the error. First, the total maximum thinning ratio of the cylindrical cup.

www.SID.ir
© 2013 IAU, Majlesi Branch
96 Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March – 2013

Table 9 Analysis of variance table for maximum thinning ratio (Parabolic cup)
Factor Degree of freedom Sum of square Mean square Contribution (%)
A 2 1.5655 0.7827 38.7545
B 2 2.0346 1.0173 50.3955
C 2 0.4175 0.2087 10.3406
*D 2 0.0197 0.0098 0.4892
Error 0
Total 8 4.0374 2.0187 100
(Error) 2 0.0790 0.0395 1.9570
*Factors used for pooling

Table 10 Analysis of variance table for maximum thinning ratio (Hemispherical cup)
Factor Degree of freedom Sum of square Mean square Contribution (%)
A
B
C
*D
2
2
2
2
8.3844
6.6811
3.3976
0.1816
4.1922
3.3405
1.6988
0.0908

I D 44.9691
35.8339
18.2227
0.9741
Error
Total
(Error)
*Factors used for pooling
0
8
2
18.6448
0.72654

f S 9.3224
0.36327
100
3.8967

Factor
A
B
Degree of freedom
2
2
o
Table 11 Analysis of variance table for maximum thinning ratio (Cylindrical cup)
Sum of square

e1.4130
0.6727
Mean square
0.7065
0.3363
Contribution (%)
37.4035
18.8078
C
D
Error
2
2
0

i v 0.3125
1.3795
0.1562
0.6897
8.2736
36.5149

Total

c
8

h 3.7779 1.8889 100

r
The percentage contribution of the selected factors and
their optimal levels are compared among the three
different formed cups, which are shown in Fig. 9 and

A Table 12, respectively. The optimal level of various

factors and their relative contributions on the maximum
thinning ratio in both parabolic cup and hemispherical
cup are almost the same while the results for cylindrical
cup are a little different. From Fig. 9, it can be inferred
that the fluid pressure is one of the first two significant
factors for all the three formed cups. As shown in Fig.
8, for the parabolic cup and hemispherical cup by
increasing the fluid pressure, the maximum thinning
ratio initially decreases, but then after a certain level
the maximum thinning ratio increases slowly again.
The high contribution of this factor is due to the fact
that increasing the fluid pressure to a certain value
increases the useful friction holding effect between the
Fig. 9 Comparison of the relative contribution of each punch and the blank during the forming process, which
factor among the three different case studies can decrease the maximum thinning ratio of the formed
cups considerably.

© 2013 IAU, Majlesi Branch www.SID.ir

Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March - 2013 97

Table 12 Comparison of the optimal level of each factor among the three different case studies
Parameter Optimum level
Parabolic cup Hemispherical cup Cylindrical cup
A: Fluid pressure 2 2 1
B: Friction coefficient 3 3 3
C: Gap (g) 1 1 1
D: Die entrance radius 3 3 3

I D
f S
e o
i v
Fig. 10

c h
Maximum principle plastic strain distribution of the formed parts (a) parabolic cup, (b) hemispherical cup, (c) cylindrical
cup

A r
By considering the results of Table 12 and Fig. 8, it can
be concluded that the amount of maximum thinning
ratio decreases with the increase of the friction
coefficient at blank/punch interface. This trend is in
punch nose, spread the strain over a greater area, move
the critical zone form the punch nose to the cup’s body
and finally reduce the maximum thinning ratio in the
formed cup. Therefore, a small increase in the amount
agreement with the observations reported in reference of friction coefficient at blank/punch interface can
[27]. Another phenomenon that can be inferred from reduce the maximum thickness reduction in dome
Fig. 9 is that the relative contribution of this factor shaped parts significantly. As it can be seen from
shows a clear decreasing trend with the change of the Tables 6 to 8, the maximum difference between the
geometry of the formed cups from parabolic cup to average S/N ratios of the friction coefficient at
cylindrical cup. As it is shown in Fig. 10, in the blank/punch interface, is |0.23849-0.27210| = 0.03361
parabolic cup and hemispherical cup, strain is for the parabolic cup which is higher than those
concentrated near the punch nose while the critical obtained for hemispherical cup (|0.10312-0.12894| =
zone of the cylindrical cup is at the cup wall near the 0.02581) and cylindrical cup (|0.11839-0.12738| =
punch profile. Thus, for dome shaped parts, specially 0.00898). This result indicates that compared to
for the parabolic cup the maximum thinning takes place hemispherical and cylindrical cups, maximum thinning
at the punch nose during the first one third of the in the parabolic cup is more sensitive to the effect of
drawing stage. Higher friction coefficient at friction coefficient at blank/punch interface, and a
blank/punch interface can reduce the tension at the greater improvement in maximum thinning can be

www.SID.ir
© 2013 IAU, Majlesi Branch
98 Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March – 2013

obtained for this cup by increasing the friction obtained through this research are summarized as
coefficient at blank/punch interface. Therefore, it can follows:
be concluded that as the geometry of punch head gets 1- For all the three case studies fluid pressure in the die
sharper, the degree of importance of the friction cavity is one of the first two significant factors. Thus,
coefficient at blank/punch interface increases the appropriate choice of this factor is very crucial for
consequently. The parameter of gap (g) has a similar minimizing thickness thinning of the formed parts and
effect on the maximum thickness reduction of all the accordingly achieving a higher drawing ratio.
three formed cups. As it is shown in Table 12 and Fig. 2- The results of ANOVA revealed that friction
8, it is obvious that the maximum thinning ratio of the coefficient at blank/punch interface is also a very
formed cups can be decreased by decreasing the important factor. The maximum thinning ratio
amount of gap (g). Based on Eq. (2), by decreasing the decreases by increasing the friction coefficient at
amount of gap (g), higher radial pressure is loaded onto blank/punch interface. The comparison of the results of
the rim of the blank which can decrease the drawing S/N and ANOVA analysis among the three formed
force and facilitate material flow into the die cavity. cups indicated that as the geometry of punch head gets
Therefore, less thickness reduction will occur in the
formed cups and accordingly the drawing ratio will
increase. Die entrance radius is found to have less
contribution on the maximum thickness reduction in
increases consequently.
D
sharper, the degree of importance of this factor

I
3- The maximum thinning ratio of the formed cups
decreases by decreasing the amount of gap (g), which
the parabolic and hemispherical cups compared to that
in the cylindrical cup. As it is shown in Fig. 9, this
factor has a negligible effect on the parabolic and
hemispherical cups, because, as illustrated in Fig. 10(a)

f S
means that increasing the radial pressure always has a
positive influence on the hydroformability of the sheet
during the HDDRP process.
4. It is shown that die entrance radius has a negligible
and (b), in these cups the critical zone is at the punch
head and hence the thickness reduction in the critical
zone is free from any bending and unbending effect.
But as shown in Fig. 10(c), the critical zone of the
cylindrical cup is at the cup wall near the punch profile
o
effect on the parabolic and hemispherical cups, while
for the cylindrical cup, it is found to be the second most
influential factor affecting the maximum thinning ratio.

e
This phenomenon can be attributed to the difference in
their critical zones.

i v
radius where bending and unbending takes place during
the forming process. This leads to larger deformation
and causes more thickness reduction in this zone.
5- Based on the results from ANOVA analysis, further
optimization of the forming parameters can be done
based on the degree of importance of the factors on the

h
Therefore, die entrance radius has a significant sheet hydroformability.
influence on the maximum thinning ratio of the

c
cylindrical cup. As it can be seen from Table 12 and
Fig. 8, maximum thickness reduction of the formed

r
cups decreases with the increase of the die entrance
radius.
REFERENCES

[1] Zhang, SH., and Danckert, J., “Development of hydro-

mechanical deep drawing”, Journal of Material

8 CONCLUSION
A
In this paper, by integrating FE method and Taguchi
technique, a systematic method has been provided to
[2]

[3]
Processing Technology, Vol. 83, 1998, pp. 14-25.
Lang, LH., Wang, ZR., Kang, DC., Yuan, S.J., Zhang,
S.H., Dankert, J., and Nielsen, K.B., “Hydroforming
highlights: sheet hydroforming

Technology, Vol. 151, 2004, pp. 165-177.

and
hydroforming”, Journal of Material Processing
tube

Kandil, A., “An experimental study of hydroforming

evaluate the effects of forming parameters on the deep drawing”, Journal of Material Processing
quality of part formability in the process of Technology, Vol. 134, 2003, pp. 70-80.
hydrodynamic deep drawing assisted by radial [4] Zhang, SH., Wang, ZR., Xu, Y., Wang, Z. T., and
pressure. Also, the analysis of variance (ANOVA) was Zhou, L. X., “Recent developments in sheet
used to quantify the influence of the forming hydroforming technology”, Journal of Material
Processing Technology, Vol. 151, 2004, pp. 237-241.
parameters on the quality characteristic. In order to [5] Thiruvarudchelvan, S., and Travis, FW., “Hydraulic-
have more comprehensive study, three different case pressure enhanced cup-drawing processes-an
studies were considered in this investigation. Hence, appraisal”, Journal of Material Processing
the results obtained from this study can be used for a Technology, Vol. 140, 2003, pp. 70-75.
wide range of industrial parts. The main conclusions [6] Zhang, SH., Zhou, LX., Wang, ZT., and Xu, Y.,
“Technology of sheet hydroforming with a movable

© 2013 IAU, Majlesi Branch www.SID.ir

Int J Advanced Design and Manufacturing Technology, Vol. 6/ No. 1/ March - 2013 99

female die”, International Journal of Machine Tools & process”, Proc IMech E B Journal of Engineering
Manufacture, Vol. 43, 2003, pp. 781-785. Manufacture, Vol. 220, 2006, pp.1937-1944.
[7] Lang, L., Danckert, J., and Nielsen, KB., [17] Singh, SK., and Kumar, DR., “Effect of process
“Investigation into hydrodynamic deep drawing parameters on product surface finish and thickness
assisted by radial pressure Part I. Experimental variation in hydro-mechanical deep drawing”, Journal
observations of the forming process of aluminum of Material Processing Technology, Vol. 204, 2008,
alloy”, Journal of Material Processing Technology, pp. 169-178.
Vol. 148, 2004, pp. 119-131. [18] Lee, S.W., “Study on the forming parameters of the
[8] Lang, L., Danckert, J., and Nielsen, KB., “Study on metal bellows”, Journal of Material Processing
hydromechanical deep drawing with uniform pressure Technology, Vol. 130-131, 2002, pp. 47-53.
onto the blank”, International Journal of Machine [19] Davidson, MJ., Balasubramanian, K., and Tagore,
Tools & Manufacture, Vol. 44, 2004, pp. 495-502. GRN., “Experimental investigation on flow-forming
[9] Gorji, A. H., Alavi, H., Bakhshi, M., Nourouzi, S., and of AA6061alloy—A Taguchi approach”, Journal of
Hosseinipour, S. J., “Investigation of hydrodynamic Material Processing Technology, Vol. 200, 2008, pp.
deep drawing for conical-cylindrical cups”, 283-287.
International Journal of Advance Manufacturing [20] He, X., Yu, Z., and Lai, X., “Robust parameters
Technology, Vol. 56, 2011, pp. 915-927. control methodology of microstructure for heavy
[10] Gorji, A. H., Alavi, H., Bakhshi, M., Valizadeh, M. E.,
and Shirkhorshidian, A., “Finite Element Simulation
and Experimental Study of Forming Conical Parts
Using Hydrodynamic Deep Drawing with Radial
[21]
D
forgings based on Taguchi method”, Journal of

I
Materials and Design, Vol. 30, 2009; pp. 2084-2089.
Roy, R., A Primer on the Taguchi Method, 1nd ed.,
New York, Van Nostrand Reinhold, 1990.

S
Pressure”, Journal of Mechanic and Aerospace- Emam [22] Lang, L., Danckert, J., and Nielsen, KB.,
Hossien Uni (in Persian), Vol. 5(3), 2009 , pp. 69-79. “Investigation into hydrodynamic deep drawing
[11] Hsu, T. C., and Hsieh, S. J., “Theoretical and assisted by radial pressure Part II. Numerical analysis

f
experimental analysis of failure for the hemisphere of the drawing mechanism and the process
punch hydroforming processes”, Journal of parameters”, Journal of Material Processing
Manufacturing Science and Engineering, Vol. 118, Technology, Vol. 166, 2005, pp. 150-161.

[12]
1996, pp. 434-438.
Zhang, SH., Jensen, MR., Nielsen, KB., Danckert, J.,
Lang, L., and Kang, D. C., “Effect of anisotropy and
prebulging on hydromechanical deep drawing of mild
steel”, Journal of Material Processing Technology,

e o
[23]

[24]
Sharma, A. K., and Rout, D. K., “Finite element
analysis of sheet Hydromechanical forming of circular
cup”, Journal of Material Processing Technology, Vol.
209, 2009, pp. 1445-1453.
Roy, RK., Design of Experiment Using Taguchi

iv
Vol. 142, 2003, pp. 544-550. Approach: 16 Steps to Product and Process
[13] Lang, L., Danckert, J., and Nielsen, KB., Improvement, 1nd ed., Wiley Inter science, New York,
“Investigation into the effect of pre-bulging during 2001.
hydromechanical deep drawing with uniform pressure [25] ABAQUS Inc, ABAQUS/Explicit User’s Manual.

h
onto the blank”, International Journal of Machine Version 6.7, 2007.
Tools & Manufacture, Vol. 44, 2004, pp. 649-657. [26] Nalbant, M., Gokkaya, H., and Sur, G., “Application
[14] Lang, L., Danckert, J., and Nielsen, KB., “Analysis of of Taguchi method in the optimization of cutting

[15]
pp. 845-856.

r c
key parameters in sheet hydroforming combined with
stretching forming and deep drawing”, Proc IMechE B
Journal of Engineering Manufacture, Vol. 218, 2004,

Lang, L., Li, T., Zhou, X., Danckert, J., and Nielsen,
[27]
parameters for surface roughness in turning”, Journal
of Materials and Design, Vol. 288, 2007, pp. 1379-
1385.
Hosseinzade, M., Mostajeran, H., Bakhshi-Jooybari,
M., Gorji, AH., Norouzi, S., and Hosseinipour, S. J.

[16]
2007, pp. 304-308. A
KB., “The effect of the key process parameters in the
innovative hydroforming on the formed parts”, Journal
of Material Processing Technology, Vol. 187-188,

Fazeli, A., and Dariani, BM., “Parameter study of the

axisymmetric hydromechanical deep drawing
“Novel combined standard hydromechanical sheet
hydroforming process”, Proc IMechE B Journal of
Engineering Manufacture, Vol. 224, 2010, pp. 447-
457..

www.SID.ir
© 2013 IAU, Majlesi Branch