You are on page 1of 6

Scientia Iranica B (2011) 18 (6), 12611266

Sharif University of Technology


Scientia Iranica
Transactions B: Mechanical Engineering
www.sciencedirect.com

Proposal of a new design for valveless micropumps


H. Afrasiab, M.R. Movahhedy , A. Assempour
Center of Excellence in Design, Robotics and Automation, Department of Mechanical Engineering, Sharif University of Technology, Tehran,
P.O. Box 11155-9567, Iran
Received 7 February 2011; revised 9 August 2011; accepted 8 October 2011

KEYWORDS
Valveless micropump;
Finite element simulation;
FluidStructure Interaction
(FSI) analysis;
Arbitrary
LagrangianEulerian (ALE)
method.

Abstract A new design for a valveless micropumping device has been proposed that integrates two
existing pumping technologies, namely, the wall induced traveling wave and the obstacle-type valveless
micropump. The liquid in the microchannel is transported by generating a traveling wave on the channel,
while the placing of two asymmetric trapezoid obstacles, along the centerline of the channel inlet
and outlet, leads to a significant (up to seven times) increase of the net flow rate of the device. The
effectiveness of this innovative design has been proved through a verified three-dimensional finite
element model. FluidStructure Interaction (FSI) analysis is performed in the framework of an Arbitrary
LagrangianEulerian (ALE) method.
2012 Sharif University of Technology. Production and hosting by Elsevier B.V.
Open access under CC BY-NC-ND license.

1. Introduction
The growing requirement of controlled fluid transfer
in many MEMS and BioMEMS applications has made the
development of microfluidic systems an area of increasing
interest in recent years. Dispensing drugs and other therapeutic
agents, fuel delivery in miniaturized fuel cells and provision of
refrigeration liquid in the cooling of microelectronic systems
are some examples of these applications.
Micropumps are essential components of microfluidic
systems. Several different types of micropump have been
developed up to now, valveless piezoelectric micropumps of
which are in wide practical use amongst which valveless
piezoelectric micropumps are in wide practical use, due to
their simple structures, high flow rate output and ability to
conduct particles such as red blood cells, polymers and proteins.
These kinds of micropump are driven by a piezoelectric element
bounded to a flexible membrane, and are equipped with a

Corresponding author.
E-mail address: movahhed@sharif.edu (M.R. Movahhedy).

1026-3098 2012 Sharif University of Technology. Production and hosting by


Elsevier B.V. Open access under CC BY-NC-ND license.
Peer review under responsibility of Sharif University of Technology.
doi:10.1016/j.scient.2011.11.023

pressure chamber for generating a pressure difference between


the inlet and outlet, and no interior moving mechanical
part is present in them. Instead, flow directing components
like nozzle/diffuser elements or asymmetric obstacles are
utilized to control rectification and obtain the net volume
of the fluid flow [1,2]. Though piezoelectric micropumps
can transport liquid at a high flow rate and with excellent
controllability, the presence of the pressure chamber makes
further miniaturization difficult and sometimes impractical [3].
Seeking a remedy for this problem, recently a mechanical
micropumping device was fabricated and presented in [3],
which uses a vibrating microchannel wall to drive the fluid
flow while no pressure chamber is present. As reported in [3],
on-generating a traveling wave on the channel wall, clear
liquid flow could be induced within the microchannel. The
transportation of the liquid by a traveling wave is controlled
by changing the magnitude and frequency of the generated
wave. A range of frequencies between 50 and 600 kHz were
considered in [3] for oscillation of the microchannel wall for a
maximum oscillation magnitude of 142.6 nm. A net outlet flow
rate of 1.20 nL/s for the vibration frequency of 100 kHz was
reported in [3] for a microchannel of 200 m width and 300 m
height. The simple structure of this Traveling Wave Micropump
(TWM) makes it advantageous for many microfluidic systems.
In this paper, we propose a simple geometric modification
to the traveling wave micropump so as to increase its net outlet
flow rate. This modification is inspired by the action mechanism
of the obstacle-type valveless micropump.
The paper is organized as follows. The working principle of
the new micropumping device is described in Section 2. The
finite element modeling procedure is presented in Section 3.
In Section 4, the results of FEM simulations are presented

1262

H. Afrasiab et al. / Scientia Iranica, Transactions B: Mechanical Engineering 18 (2011) 12611266

Figure 1: Working principle of the obstacle-type valveless micropump [5].

micropump. Now, we place two trapezoid obstacles along


the centerline of the microchannel, one in the channel inlet
downstream of the electrodes array and the other in the channel
outlet upstream of the electrodes array, as shown in Figure 2.
These obstacles form two equivalent planar nozzle/diffuser
rectifying elements, similar to those described for the obstacletype valveless micropumps above. Consequently, a directiondependent flow resistance is provided, and an increase in the
net flow rate of the microchannel is expected based on the
previous experience with obstacle-type micropumps. In the
next sections, we examine the effect of the proposed geometric
modification on the net outlet flow rate of the microchannel,
using a finite element simulation of both with- and withoutobstacle microchannels.
3. Finite element method simulation

Figure 2: A schematic of the proposed micropump.

and discussed. Finally, the concluding remarks are reviewed in


Section 5.
2. Working principle of the new micropump
In order to shed light on the action mechanism of the
proposed micropump, first we briefly review the working
principle of the obstacle-type valveless micropump presented
e.g. in [4,5]. This kind of micropump is composed of a pump
chamber with an oscillating diaphragm and two nozzle/diffuser
flow rectifying elements, one directed from the inlet to the
pump chamber, and the other from the pump chamber to
the outlet, as illustrated in Figure 1. Each element is formed
by two equivalent regions between the symmetrical trapezoid
obstacle and the side-walls of the microchannel. Actuating the
diaphragm causes a continual periodic increase and decrease
in the volume of the pumping chamber (Figure 1). When the
volume of the pumping chamber increases, the pressure in
the chamber decreases, and more fluid enters through the
nozzle-diffuser element on the left (inlet) relative to that on the
right (outlet). This is because the element on the left acts as a
diffuser, which poses less flow resistance than the nozzle on the
right. Conversely, when the volume of the pumping chamber
decreases, more fluid exits through the element on the right,
which now acts as a diffuser. This results in a net pumping
action from left to right, in Figure 1, as the diaphragm vibrates
up and down.
Now, we turn our attention to the new micropump. A
schematic of this device is illustrated in Figure 2. In this device,
the fluid flow is driven by the vibrating microchannel wall
and actuated by a piezoelectric layer. The top ceiling wall
of the microchannel is composed of a silicon wafer whose
surface is covered by a piezoelectric PZT thin film. Nine separate
top electrodes are attached to the PZT film. The traveling
wave is induced on the ceiling wall by applying sinusoidal
voltages, with a phase difference of 2 /3, to the adjacent
top electrodes. When a channel wall oscillates in the form
of traveling waves, peristaltic motion is induced in the liquid
beneath the microchannel wall [3]. After a period of oscillation,
the fluid moves slightly forward from the initial position, due to
its viscosity, and the net liquid can be transported by repeating
the period of motion. Up to this point, the pumping principle
of the device is exactly similar to that of the traveling wave

In order to investigate the effect of obstacle addition on


the net flow rate of the micropump, numerical simulations of
both with- and without-obstacle microchannels were carried
out in this section, using the finite element method. As stated
earlier, the fluid flow is driven by the vibrating microchannel
wall. Since the fluid also plays a role in resisting the wall
vibration, the solid wall vibration and fluid flow are coupled,
and a full FluidStructure Interaction (FSI) analysis is required
in order to investigate the performance characteristics of these
devices [6,7]. For this purpose, the FSI analysis is performed,
using an Arbitrary LagrangianEulerian (ALE) approach in the
framework of the finite element method. The simulations are
performed by means of a verified in-house developed FSIALE code, which has been used previously, e.g. in [8,9]. The
finite element modeling details are described in the following
subsections.
3.1. The fluid flow governing equations
The law of conservation of momentum and the continuity
equation for a viscous Newtonian incompressible flow are
NavierStokes equations. In the ALE approach to fluidstructure
interaction problems, these equations are written in the ALE
description to account for deformation of the fluid mesh in the
FSI interface. These equations, along with appropriate initial
and boundary conditions, are seen in the ALE description (see
e.g. [10])

v
2f . s v + v vm . v + p

t
= bf

in f (0, T ) ,

(1a)

.v = 0 in f (0, T ) ,

(1b)

v (x, 0) = v0

(1c)

on {0} ,
f

v (x, t ) = vD (x, t )

on

f
D

(0, T ) ,

n . = pn + 2f n . v = t
f

on

(1d)
f
N

(0, T ) ,

(1e)

where f , f , v and p represent the fluid density, viscosity,


velocity and pressure, respectively. vm is the mesh velocity and
bf denotes the body force vector. The fluid domain is specified
by f , while the time interval of interest is between (0, T ). The
()

time derivative in the ALE frame is represented by t . The

symmetric tensor, s v = v + ( v)T /2, is called the rate


of deformation (or strain rate) tensor. nf is the exterior normal
to the fluid boundary, f , and v0 is the initial condition for

H. Afrasiab et al. / Scientia Iranica, Transactions B: Mechanical Engineering 18 (2011) 12611266

1263

the velocity field. vD denotes the prescribed velocity on the D


portion of the boundary, and tf is the boundary traction on the
f
complementary portion, N .
Since the working frequency of the solid wall is very high
(between 50 and 600 kHz), the time step size required for the
finite element analysis of these micropumps is very small. Thus,
in order to avoid the so-called small time step instability in the
finite element analysis of the fluid part, we used a residualbased variational multiscale formulation for incompressible
NavierStokes equations, in combination with stabilization
parameters, as recommended in [11], for weak formulation of
the fluid flow.
3.2. The solid governing equations
In a standard Lagrangian description, the law of conservation
of linear momentum for a solid continuum may be expressed
with respect to spatial coordinate, x, as:

2u
= .s + s bs in s (0, T ) ,
t2
u (x, 0) = u0 (x) in s {0} ,
u
(x, 0) = u 0 (x) in s {0} ,
t
u (x, t ) = uD (x, t ) in Ds (0, T ) ,
s

(2a)

Figure 3: The sequential fluidstructure coupling algorithm.

f vm . v takes the convection of mesh moving points into

(2d)

account. The solid structure is formulated and solved in the


Lagrangian description, as indicated earlier in this paper.
Different techniques of ALE mesh moving are presented
in the literature. Here, we use an approach that is based on
solving the linear elasticity equations [12]. In this approach, the
equation governing the displacement of fluid mesh nodes can
be written as:

.n = t on
(0, T ) ,
(2e)
where s is the solid density, and u represents its displacement

. = 0 in f (0, T ) ,
(5)
where is the Cauchy stress tensor. For each boundary, i, a

(2b)
(2c)

Ns

field, whereas the body forces are given by vector bs . The


symmetric second order tensor, s , denotes the Cauchy stress
tensor, and the solid domain is represented by s . uD denotes
the prescribed velocity on the Ds portion of the boundary, and ts
is the boundary traction on the complementary portion, Ns . The
outer unit normal to the solid boundary is specified by vector ns .
Since the solid membrane deformation is small in most
microfluid manipulating devices [8], we assume that the
structural part of the FSI problem is governed by linear
elasticity. For a linear elastic solid, we have:

s = s (.u) I + 2s (u) ,
(3)

T
in which (u) = u + ( u) /2 is the strain tensor, and s
and s are the Lam coefficients.
3.3. Fluidstructure interface conditions
Fluidsolid interface conditions consist of kinematic and
dynamic constraints specified as follows on the FSI interface,
fs :

us
= vf
t

on fs (0, T ) ,

s .n + f .n = 0 on fs (0, T ) ,

(4a)
(4b)

with n being the outer normal at the solid boundary.


3.4. The mesh motion algorithm
In the Arbitrary LagrangianEulerian (ALE) approach to
fluidstructure interaction problems, two key ingredients are
needed. The first is a technique to move the fluid mesh, so
that it can track structure motion at the FSI interface. This
deformation is extended into the fluid interior field, in order
to avoid excessive distortion of the fluid mesh at the interface.
The second ingredient is a flow formulation written in the
ALE framework, as presented in Eq. (1). The time derivative in
Eq. (1a) is taken within the ALE description and the term,

Dirichlet boundary condition:


ui = ubi ,

(6)
b

may be given. u is the boundary displacement vector that


is either given a priori or computed by the solid structure
equations.
3.5. Fluidstructure coupling method
In this work, a partitioned strong coupling approach
is employed that uses a sequential iterative algorithm, as
illustrated in Figure 3. This iterative scheme for the coupled
system of fluid and structure is repeated at each time step until
convergence is reached, i.e. after the norms of relative change
in the field variables between two consecutive iterations are
smaller than the specified convergence tolerance.
3.6. Boundary and initial conditions
The boundary conditions are as follows: Along the lateral
and bottom walls of the microchannel and obstacles (fixed
boundaries), the usual no slip boundary condition is prescribed for the fluid. Along the top wall of the microchannel,
which is also the bottom of the microchannel ceiling (FSI boundary), the fluidstructure interface conditions given in Sections 3
and 4 are applied. The motion of the microchannel ceiling is
prescribed on its upper face. The initial condition is zero flow
velocity and structure (microchannel ceiling) displacement.
3.7. Verification of the numerical approach
In order to verify the accuracy of the finite element approach,
a portion of the without-obstacle microchannel (containing all
electrodes, as well as 1 mm downstream and 1 mm upstream
of the electrodes array) was simulated in three dimensions, for
20 complete periods of oscillation and for different values of
oscillation frequency. 14 040 and 3120 tri-quadratic elements

1264

H. Afrasiab et al. / Scientia Iranica, Transactions B: Mechanical Engineering 18 (2011) 12611266


Table 1: Without-obstacle microchannel geometry.
Silicon wafer thickness (m)
Piezoelectric film thickness (m)
Distance between single electrodes centers (m)
Excitation voltage (V)
Maximum oscillation amplitude of electrodes for
frequency of 100 kHz (nm)
Microchannel width (m)
Microchannel height (m)
Microchannel length (mm)

Table 3: With-obstacle microchannel geometry.


6.0
2.5
400
20
142.6

Distance between two obstacles (L) (mm)


Obstacles length (L1 ) (m)
Microchannel width (W ) (m)
Obstacles width (W1 ) (m)
Obstacles divergence angle ( )

3.2
500
200
25
5

200
300
5.2

Table 2: Solid physical properties.


Material
Silicon
Piezoelectric

Density (kg/m3 )
2300
7800

Elastic
modulus (GPa)
170
64

Poissons
ratio

Figure 5: A schematic top view of the micropump geometry.

0.215
0.31

Figure 4: Frequency dependence of the mean flow velocity.

were used to discretize the fluid and solid parts, respectively.


We used 80 time steps per period of oscillation in all these
simulations. The without-obstacle microchannel geometry is
presented in Table 1. The working fluid is water with density of
1000 kg/m3 and viscosity of 0.001 kg/(ms). The solid physical
properties are listed in Table 2.
In each case, the mean flow velocity was calculated in the
microchannel outlet. Figure 4 compares the results of these
simulations with the experimental data reported in [3]. It can
be seen that the finite element results are in good agreement
with experimental data, which demonstrates the capability of
the employed numerical model.
4. Results and discussion
Once the numerical model is verified, it can be used
for studying the performance characteristics of the proposed
micropump. To this end, in this section, a simulation similar
to that of Section 3 was carried out for the with-obstacle
micropump, for wall oscillation frequency of 100 kHz. 16 448
and 4112 tri-quadratic elements were used to discretize the
fluid and solid parts, respectively. A schematic top view of the
micropump geometry is indicated in Figure 5, and detailed
dimensions of the obstacles and microchannel are listed in
Table 3. The dimensions of the obstacles are adopted from
those presented in [4]. Other geometric dimensions of the withobstacle micropump that are not given in this table are the
same as those of the without-obstacle micropump presented
in Table 1. It is to be noted that due to the presence of the

Figure 6: Outlet flow rate variation for without- and with-obstacle micropumps.

obstacles, more energy is needed for suction and discharge


of fluid into and from the micropump. Consequently, in the
new micropump, a higher voltage is needed for deforming
the microchannel wall that controls the fluid flow. An electrofluidstructural interaction simulation (see e.g. [13,14]) is
needed for determination of the required increase in the
operating voltage.
It is shown in [8] that after approximately five pumping
cycles, the flow rate of the TWM micropump (without obstacle)
finds a harmonic variation, which is in accordance with the
harmonic deformation of the microchannel wall. The current
study indicates that the same matter also holds for the
proposed micropump (with obstacle). In other words, after
approximately 5 cycles, the flow rate is the same for every cycle
in the case of the TWM micropump, and for every other cycle
in the case of the proposed micropump. Consequently, using
the results of four cycles (which we selected to be the 7th to
10th cycles) seems to be adequate for comparing the flow rate
of the two micropumps. The variation of the flow rate at the
outlet of with- and without-obstacle micropumps is presented
in Figure 6 for the 7th to 10th cycle of pumping. The first half
of each cycle, where the flow rate curve is positive, corresponds
to the pump phase in which fluid exits the microchannel from
both the inlet and outlet, while in the second half of each cycle,
the flow rate curve is negative, and fluid enters from both ends
of the microchannel to accomplish the supply phase. The cyclic
average of the outlet flow rate of each micropump is obtained by
taking the average of its respective outlet flow rate (the curves

H. Afrasiab et al. / Scientia Iranica, Transactions B: Mechanical Engineering 18 (2011) 12611266

1265

Figure 7: The cyclic average of outlet flow rates of micropumps.

Figure 9: The fluid pressure contour for the with-obstacle micropump (in Pa).

Table 4: Cyclic average of the outlet flow rate of the 7th pumping cycle for
different levels of element numbers.
Number of elements
6 410
10 240
14 720
20 560
32 660

Figure 8: The fluid pressure contour for the without-obstacle micropump


(in Pa).

of Figure 6) over one cycle. The averaging results for withoutand with-obstacle micropumps are presented in Figure 7 for
seventh to tenth cycles. It can be seen from this figure that
adding obstacles increases the cyclic average flow rate of the
micropump from around 1 nL/s to around 6.5 nL/s, which is
quite significant. This proves the efficiency of the proposed
geometrical modification in increasing the net outlet flow rate
of the Traveling Wave Micropump (TWM).
The role of obstacles in the proposed micropump is to
provide directional flow resistance (as explained in Section 2),
which causes a better performance in terms of one-directional
flow (from inlet to outlet). For example, in the 7th cycle,
the total volume of fluid moved (i.e. back and forth) is
2.59 L/s for the without-obstacle and 2.35 L/s for the
with- obstacle micropump. However, the cyclic average of the
outlet flow rate, which is a measure of the total volume of
fluid moved in the desired direction (from inlet to outlet), is
1.03 L/s for the without-obstacle and 6.79 L/s for the withobstacle micropump. It can be seen that in the without-obstacle

The outlet flow rate (nL/S)


5.48
6.07
6.62
6.79
6.79

micropump, a higher total volume of fluid is moved, but due


to the presence of obstacles, the outlet flow rate of the with
obstacle micropump is significantly higher.
The FEM result for the cyclic average of the outlet flow rate of
the 7th pumping cycle is presented in Table 4, for five different
levels of element number. As this table indicates, discretizing
the problem domain by 20 560 elements is sufficient for
convergence of the simulations. A similar examination in time
showed that a time step size of 1t = 1.25 107 (80 time
steps per period of microchannel wall oscillation) suffices for
obtaining a convergent result.
The three-dimensional contour plots of the microchannel
fluid pressure at different instances of the tenth cycle, for
without- and with-obstacle micropumps, are presented in
Figures 8 and 9, respectively.
The maximum pressure difference generated in the microchannels of without- and with-obstacle micropumps is compared in Figure 10, for different instances of the tenth cycle.
According to this figure, the maximum pressure difference is
greater for with-obstacle micropump at all time instances.
5. Conclusion
In this paper, a simple geometric modification to the
traveling wave micropump has been proposed by placing
two asymmetric trapezoid obstacles along the centerline of
the micropump inlet and outlet. This modification is inspired
by the action mechanism of the obstacle-type valveless
micropump. The effect of this modification is examined
through a verified finite element simulation of the proposed
micropump. Fluidstructure interaction is considered in the
simulation, and the variational multiscale method is used for
fluid formulation in the ALE description. It is shown that adding

1266

H. Afrasiab et al. / Scientia Iranica, Transactions B: Mechanical Engineering 18 (2011) 12611266

Figure 10: The maximum pressure difference generated in the microchannels.

obstacles leads to approximately seven times increase in the


outlet flow rate of the traveling wave micropump.
References
[1] Stemme, E. and Stemme, G. A valveless diffuser/nozzle-based fluid
pump, Sensors Actuators A, 39, pp. 159167 (1993).
[2] Koch, M., Harris, N., Evans, A.G.R., White, N.M. and Brunnschweiler, A. A
novel micromachined pump based on thick-film piezoelectric actuation,
Sensors Actuators A, 70, pp. 98103 (1998).
[3] Ogawa, J., Kanno, I., Kotera, H., Wasa, K. and Suzuki, T. Development
of liquid pumping devices using vibrating microchannel walls, Sensors
Actuators A, 152, pp. 211218 (2009).
[4] Lee, C.J., Sheen, H.J., Tu, Z.K., Lei, U. and Yang, C.Y. A study of PZT
valveless micropump with asymmetric obstacles, Microsyst. Technol., 15,
pp. 9931000 (2009).
[5] Sheen, H.J., Hsu, C.J., Wu, T.H., Chang, C.C., Chu, H.C., Yang, C.Y. and Lei, U.
Unsteady flow behaviors in an obstacle-type valveless micropump by
micro-PIV, Microfluid. Nanofluid., 4, pp. 331342 (2008).
[6] Pan, L.S., Ng, T.Y., Liu, G.R., Lam, K.Y. and Jiang, T.Y. Analytical solutions
for the dynamic analysis of a valveless micropumpa fluid-membrane
coupling study, Sensors Actuators A, 93, pp. 173181 (2001).
[7] Zhou, Y. and Amirouche, F. Study of fluid damping effects on resonant
frequency of an electromagnetically actuated valveless micropump, Int. J.
Adv. Manuf. Technol., 45, pp. 11871196 (2009).

[8] Afrasiab, H., Movahhedy, M.R. and Assempour, A. Fluidstructure


interaction analysis in microfluidic devices: a dimensionless finite
element approach, Int. J. Numer. Methods Fluids, doi:10.1002/fld.2592
(2011).
[9] Afrasiab, H., Movahhedy, M.R. and Assempour, A. Finite element
and analytical fluidstructure interaction analysis of the pneumatically
actuated diaphragm microvalves, Acta Mech., doi:10.1007/s00707-0110508-9 (2011).
[10] Donea, J. and Huerta, A., Finite Element Methods for Flow Problems, John
Wiley & Sons, Chichester (2003).
[11] Hsu, M.C., Bazilevs, Y., Calo, V.M., Tezduyar, T.E. and Hughes, T.J.R.
Improving stability of stabilized and multiscale formulations in flow
simulations at small time steps, Comput. Methods Appl. Mech. Engrg., 199,
pp. 828840 (2010).
[12] Stein, K., Tezduyar, T.E. and Benney, R. Mesh moving techniques for
fluidstructure interactions with large displacements, J. Appl. Mech., 70,
pp. 5963 (2003).
[13] Ha, D.H., Phan, V.P., Goo, N.S. and Han, C.H. Three-dimensional electrofluidstructural interaction simulation for pumping performance evaluation of a valveless micropump, Smart Mater. Struct., 18, pp. 104111
(2009).
[14] Fan, B., Song, G. and Hussain, F. Simulation of a piezoelectrically actuated
valveless micropump, Smart Mater. Struct., 14, pp. 400405 (2005).

Hamed Afrasiab received his B.S. and M.S. degrees in Mechanical Engineering
from Sharif University of Technology, Iran in 2004 and 2006, respectively, and
is now a Doctoral degree student of the same subject at the same university.
His main research area includes the finite element modeling of fluidstructure
interaction in micromechanical systems.
Mohammad Reza Movahhedy received his B.S. degree from University of
Tehran, Iran in 1988, his M.S. degree from the University of Waterloo, Canada
in 1994, and his Ph.D. degree from the University of British Columbia, Canada in
2000, all in Mechanical Engineering. He is currently Professor in the Department
of Mechanical Engineering at Sharif University of Technology, Iran. His research
interests are FEM simulation of metal cutting/forming processes, machine tool
dynamics, mechanics of machining processes, experimental modal analysis and
computer aided tolerancing.
Ahmad Assempour received his B.S. and M.S. degrees from Tehran Polytechnic,
Iran in 1979 and 1985, respectively, and his Ph.D. degree from Oklahoma
State University, USA in 1989, all in Mechanical Engineering. He is currently
Professor in the Department of Mechanical Engineering at Sharif University of
Technology, Iran. His research interests are plasticity and mechanics of metal
forming, modeling of sheet and bulk forming and die design in auto stamping
industries.

You might also like