Design Optimization of Flexure Hinges Profile Using FEM and Experimental
Verification in Piezoelectric Actuators
H. Esteki 1, A. Shahidi3, S. Pirouzpanah2 and S. Sarkar 3
1
School of Mechanical Engineering, Amirkabir University of Technology, Tehran, Iran.
2
School of Mechanical Engineering, Sharif University of Technology, Tehran, Iran.
3
Research Center for Science and Tech. in Medicine, Imam Khomeini Hospital, Tehran, Iran.
H_esteki@cic.aut.ac.ir
Abstract: In this paper, flexure hinges as the main part of
the actuators frame are studied considering different
profiles of flexure hinges and effective design parameters to
introduce the best profile of the hinge based on the
application that is needed.
For this purpose five different profiles (right circular,
elliptical, corner-filleted, hyperbolic and parabolic) of
flexure hinges have been considered and further studies
have been carried out on flexure hinges based on some
parameters such as accuracy, flexibility and the maximum
stress in the hinge.
The amount of flexibility of the hinge determined based
on strain energy, accuracy of hinge (amount of deviation
from actual hinge) and stresses in hinge is achieved
invoking FEM analysis, we compare different profiles used
in flexure hinges and the behavior of accuracy is studied
with deflection variation.
Different flexure hinge profiles are used in a sample
mechanism and amplification factor, natural frequency and
accuracy of results are compared with FEM results.
In order to verify the obtained results, the sample
mechanism with corner-filleted flexure hinge with the fillet
radius of 0.1 mm is manufactured and the FEM results is
verified with measured displacement of this sample frame.
Keywords: flexure hinges, flexure profile, accuracy,
flexibility, amplification factor, FEM
1. INTRODUCTION
Nowadays, usage of piezoelectric actuators has been
increased significantly in precision positioning industries
such as precision X-Y stages (Moriyama et al., 1988),
piezoelectric motors (Ishida, 1992), active dampers, high-
accuracy alignment devices for optical fibers (Galvagni,
1990), scanning tunneling microscopes, high-precision
cameras, robotic micro-displacement mechanisms, antennas
and valves. Because of the low displacement of piezoelectric
actuators (only about 0.1% of its length), flexure hinges are
used to magnify the piezostacks displacement amplitude
considering their advantages such as designs requiring one-
piece(monolithic) manufacturing, reduced weight,
smoothness of movement, zero backlashes, no need of
lubrication, and virtually infinite resolution.
Paros and Weisbord (1965), in their fundamental work,
presented the design equations, both exact and simplified,
for calculating the compliances (spring rates) of single-axis
and two-axis circular cutout constant cross-section flexure
hinges. Ragulskis et al. (1989) applied the static finite
element analysis to one-quarter of circular flexure hinges in
order to calculate their compliances. The analysis results
were further used to formulate an existence criterion based
on the deformation of an initially straight cross-section; this
allowed specifying optimal flexure geometry for minimum
bending stiffness. Lobontiu et al. (2001) developed an
analytical model of corner-filleted flexure hinges that are
incorporated into planar amplification mechanisms.
Compliance factors were formulated that allow evaluating
the rotation efficiency, precision of motion and stresses. A
corner-filleted flexure hinge spans a design space that is
limited by the right circular flexure and the simple beam, in
terms of its compliance. Xu and King (1996) performed
static finite element analysis of circular, corner-filleted and
elliptic flexure hinges. The results revealed that the corner-
filleted flexure is the most accurate in terms of motion, the
elliptic flexure has less stress for the same displacement,
while the right circular flexure is the stiffest. Ryu and
Gweon (1997) modeled the motion errors that are induced by
machining imperfections into a flexure hinge mechanism.
Lobontiu et al. (2002) introduced the parabolic and
hyperbolic flexure hinges for planar mechanisms, by using
compliance closed-form solutions to characterize their
performance in terms of flexibility, precision of motion and
stresses.
At this work, to achieve the optimum amount of the
accuracy or precision of rotation quantified by the offset of a
flexures symmetry center, the flexibility quantified by strain
energy and stress in elliptic flexure hinges, an FEM analysis
is performed in terms of two non-dimensional parameters,
which determine our geometry. Also we compared the
accuracy of elliptic, parabolic, hyperbolic and circular
flexure hinges by configuring the diagram including offset of
flexures symmetry center vs. flexure's end deflection. The
mentioned flexure hinge profiles are used in a sample
mechanism and amplification factor, natural frequency and
accuracy of results are compared invoking FEM results.
In order to verify the obtained results, the sample
mechanism with corner-filleted flexure hinge with the fillet Figure 2. Cross-sectional profile of a symmetric corner-filleted flexure hinge
radius of 0.1 mm is manufactured and the FEM results is
verified with measured displacement of this sample frame. Corner-Filleted Flexure Hinge:

[ ]
t + 2 r - x(2r - x ) , x [0, r ]

2. BASIC ASSUMPTIONS t ( x ) = t , x [r , l - r ]
(5)
The geometry equations of different types of flexure hinge

{
t + 2 r - (l - x )[2r - (l - x )] } , x [l - r , r ]
considering Figure1 and Figure 2 are:
Using these equations allows us to have a parametric
model in FEM software, which provides us rapid change of
flexure hinge types and geometric parameters. The
geometric parameters are mentioned based on two non-
dimensional parameters b and g , that are defined as:
t
b=
2c
(6)
t
g =
l

3. ACCURACY, FLEXIBILITY AND STRESS STUDY

Figure 1: Parameters defining a symmetric conic-section flexure hinge
a. Accuracy
Right circular flexure hinge:
The relative rotation of two mechanical members that are
[ ]
t ( x ) =t + 2 r - x (2r - x ) ; r = c , l = 2c (1)
connected by a conventional rotation joint is produced along
an axis that passes through the geometric center of the joint,
which is fixed, provided one member is also fixed. In the
Parabolic Flexure Hinge: case of a symmetric flexure hinge, the center of rotation (the
geometric symmetry center of the flexure) is no longer fixed
2
x since the flexure is subjected to deformation.
t (x ) = t + 2c1 - 2 (2)
l In order to study the accuracy of flexure hinge, an FEM
model is created in which the material used is steel
Hyperbolic Flexure Hinge: (E=200E9,n = 0.3 ).(Figure 3) The displacement of the
flexure hinge rotation center (point 2 in Figure 1) produced
2 by input displacement of 0.1mm applied in right edge,
x
t (x ) = t 2 + 4c (c + t )1 - 2 (3) indicates the accuracy of flexure hinge. Whatever this
l displacement was smaller the flexure hinge operates closer
to a real hinge. An elliptic flexure hinge with varying t / L
Elliptical Flexure Hinge:
and t / 2c parameters ( L : constant) is studied according to
Figure 4.
2 x
2

t ( x ) = t + 2c 1 - 1 - 1 - (4)
l

Figure 3: Parametric design of an elliptic flexure hinge

 flexure profiles

From Figure 4, it can be seen that the accuracy of an

t/L elliptic flexure hinge decreases by increasing t / L
and t / 2c .
Figure 5 and 6 show that hyperbolic flexure hinge is the
most accurate profile and corner filleted flexure is the least
accurate.
Offset(m)

b. Flexibility

By calculating strain energy in FEM analysis, the

flexibility of flexure hinge can be studied. The more flexible
flexure hinge has the lower level of strain energy.

5.00E-02

4.50E-02

4.00E-02
t/2c 3.50E-02
Figure 4: Geometric center's offset in an elliptical flexure hinge
Strain Energy

3.00E-02

Different types of flexure with the same t / L and 2.50E-02

2.00E-02
t / 2c are analyzed in FEM software. The results are shown 1.50E-02
in Figure 5 and Figure 6. 1.00E-02

(t =1, L=10, c=5 mm) 5.00E-03 0.4

0.25
0.00E+00
3.00E-05 0.18
t/2c=0.05

0.12
t/2c=0.06
t/2c=0.07

t/L
t/2c=0.08

0.09
t/2c=0.1

2.50E-05
t/2c=0.12
t/2c=0.15

0.07
t/2c=0.2
t/2c=0.25

0.05
t/2c=0.3

2.00E-05
t/2c=0.4

circular
Offset(m)

elliptic
1.50E-05
hyperbolic

t/L
parabolic
Figure 7: Level of strain energy for varying geometric parameters
1.00E-05
and t / 2c .
5.00E-06

Figure7 shows that bigger energy is consumed for

0.00E+00 deflection of flexure hinge with increasing in t / L and
0 0.2 0.4 0.6 0.8 1 1.2

Deflection(mm) decreasing in t / 2c . It means that the flexure hinge with

smaller t / L and larger t / 2c has more flexibility.
(t =1, L=10, c=r=2 mm)

7.00E-05
c. Stress Study
6.00E-05

The maximum von Mises stress, under specific

5.00E-05 constraint, which is happened in an elliptical flexure hinge,
is showed in Figure8.
4.00E-05 Corner Filleted
Figure8 shows that the maximum von Mises stress
Offset(m)

elliptic

3.00E-05
hyperbolic
parabolic
produced in an elliptical flexure hinge decreases using a
flexure with smaller t / L and larger t / 2c .
2.00E-05
It means that the flexure hinge with smaller t / L and
1.00E-05 larger t / 2c has longer fatigue life.

0.00E+00
0 0.2 0.4 0.6 0.8 1 1.2

Deflection(mm)

Figure 5 and Figure 6: Geometric center's offset vs. end deflection for different
 applications that work in low frequencies.
9.00E+08
In addition, it is apparent that the microactuator that uses
8.00E+08
hyperbolic flexure hinge is the stiffest structure.
7.00E+08

6.00E+08
Stress(Pa)

5.00E+08 5. EXPERIMENTAL VERIFICATION

4.00E+08

3.00E+08
The sample mechanism with corner-filleted flexure hinge
2.00E+08
with the fillet radius of 0.1 mm is manufactured according to
1.00E+08
Figure10. The vertical displacement of microactuator end-
0.00E+00
effecter is measured by a comparator. The experimental

t/L=0.4
0.04

t/L=0.25
results are shown in Table2.
0.06

t/L=0.18
0.08

t/L=0.12
0.1

t/L=0.09
0.15

0.25

t/L=0.07

t/2c
0.35

t/L=0.05

Figure 8: The maximum von Mises stress for varying geometric parameters
t/L and t / 2c .

4. MICRACTUATOR

A sample mechanism for microactuator is modeled in

FEM software according to Figure9 with the flexure hinge
geometric specifications of t = 0.3 , c = 0.1 , l = 1.5 mm

Figure 10: A microactuator frame test under comparator

Table 2: End-effecter displacement of microactuator frame

ANSYS Experimental Amplification Ratio
Input
Output Output ANSYS Experimental
1 4.8614E-6 4.2E-06 4.86144 4.20
2 9.7229E-6 9.1E-06 4.86144 4.55
3 1.4584E-5 1.4E-05 4.86144 4.67
4 1.9445E-5 1.87E-05 4.86144 4.68
5 2.4307E-5 2.38E-05 4.86144 4.76

As it's shown in Table2 the experimental results are close

Figure 9: Parametric design of a flexure based microactuator
to the FEM results. The differences between these results
can be interpreted by manufacturing error and presence of an
The amplification ratio and natural frequency of
external force in experimental test.
microactuator using different types of flexure hinge are in
Table1

Table 1: Microactuator comparison based on the type of flexure hinge 6. CONCLUSION

Amplification Natural
Flexure Type Hyperbolic flexure hinge is the most accurate profile and
Ratio Frequency(Hz)
Corner Filleted 4.861448 218.79 corner filleted flexure is the least accurate.
elliptic 4.854049 237.27 A flexure hinge with smaller t / L and larger t / 2c has
parabolic 4.850139 244.75 more flexibility.
hyperbolic 4.848669 247.20 A flexure hinge with smaller t / L and larger t / 2c has
longer fatigue life.
The results in Table1 indicate that the microactuator that A microactuator that uses corner filleted flexure hinge has
uses corner filleted flexure hinge has the largest the largest amplification ratio and lowest natural frequency.
amplification ratio and lowest natural frequency. This means A microactuator that uses hyperbolic flexure hinge is the
that such an actuator is suitable for displacement oriented stiffest structure.
 REFERENCES

[1] Moriyama, S., Harada, T. and Takanashi, A., "Precision

X-Y stage with a piezo-driven fine-table", Bull Japan Soc
Prec Eng, Vol.22, 1988, pp. 13-14.
[2] Ishida, M., Hamaguchi, J., Shirasuka, K. and Hori, T.,
"A new friction-type piezoelectric motor utilizing
mechanism of the strain wave gearing", IEEE Trans Ind
Electronics, Vol.39, 1992, pp. 30-35.
[3] Galvagni, J., "Electrostrictive actuators and their use in
optical applications", Opt Eng, Vol.29, 1990, pp. 1389-1391
[4] Paros, J.M., Weisbord, L., "How to design flexure
hinges", Machine Design, Vol.25, 1965, pp. 151156.
[5] Ragulskis, K.M., Arutunian, M.G., Kochikian, A.V.,
Pogosian, M.Z., "A study of fillet type flexure hinges and
their optimal design, Vibration Engineering" , Vol.3, 1989,
pp. 447-452.
[6] Lobontiu, N., Paine, J.S.N., Garcia, E., Goldfarb, M.,
"Corner filleted flexure hinges", ASME Journal of
Mechanical Design, Vol.123, 2001, pp. 346352.
[7] Xu, W., King, T.G., "Flexure hinges for piezo-actuator
displacement amplifiers: flexibility, accuracy and stress
considerations", Precision Engineering, Vol.19, 1996, pp. 4
10.
[8] Ryu, J.W., Gweon, D.G., "Error analysis of a flexure
hinge mechanism induced by machining imperfection",
Precision Engineering, Vol.21, 1997, pp. 8389.
[9] Lobontiu, N., Paine, J.S.N., Garcia, E., Goldfarb,
M.,"Design of symmetric conic-section flexure hinges based
on closed-form compliance equations", Mechanism and
Machine Theory, Vol.37, 2002, pp. 477498.