wan2016548gfdruoohfrewwwryuopjbvfddss.pdf

Design and analysis of a new compliant XY micropositioning
stage based on Roberts mechanism
Sicong Wan, Qingsong Xu ⁎
Department of Electromechanical Engineering, Faculty of Science and Technology, University of Macau, Avenida da Universidade, Taipa, Macau, China
a r t i c l e i n f o a b s t r a c t
Article history:
Received 13 January 2015
Received in revised form 5 September 2015
Accepted 13 September 2015
Available online xxxx
The demand for a compliant micropositioning stage with a large workspace and high precision is
increasing. This paper presents a new design of XY micropositioning stage based on Roberts mech-
anism, which is a traditional linear guiding mechanism. By extending the traditional rigid-body
Roberts mechanism, two kinds of new compliant guiding mechanisms are designed and a novel
parallel-kinematic XY compliant guiding stage is synthesized. Pseudo-rigid-body model (PRBM)
is developed to build the quantitative models of the compliant Roberts mechanisms. In addition,
finite-element analysis (FEA) is carried out to validate its performance. A prototype is fabricated
and tested through experimental studies. Results show that the XY stage delivers a work range
larger than 12 mm in each axis. Moreover, the parasitic motion of the stage in the non-working
direction is less than 1.7% of the motion stroke. It indicates that the designed stage meets the re-
quirements of a large stroke and high precision linear guiding mechanism, which demonstrates
the feasibility of the proposed design ideas using the Roberts mechanism.
© 2015 Elsevier Ltd. All rights reserved.
Keywords:
Compliant mechanism
XY stage
Pseudo rigid body model (PRBM)
Roberts mechanisms
Micropositioning
1. Introduction
Compliant mechanism can transmit motion and force by using the material's elastic deformation capability. As compared with
rigid-body mechanism, compliant mechanism eliminates the problems such as clearance, friction, and wears [1]. However, the appli-
cations of compliant mechanism are limited because it is hard to make a precision micropositioning system with both a large
workspace and multi-degree of freedom (DOF) at the same time [2–5]. For instance, many applications require a compliant mecha-
nism which has a larger range of work, e.g., the micro-coordinate measuring machines, adaptive adjustment mechanism in astronom-
ical telescope, MEMS processing systems, binary optical processing equipment, and so on [6,7]. Hence, the large-stroke and multi-DOF
micropositioning stage has drawn extensive attentions of some recent research. For example, Tan et al. [8] proposed a multi-leaf con-
figuration of expanded-motion-range flexure hinges. Zhao et al. [9] designed a new large-deflection pivot based on curved flexure el-
ement. Xu [10] designed a novel compact long-stroke precision positioning stage based on flexible leaf springs. Trease et al. [11]
proposed a large-displacement compliant joint, and Choi et al. [12] designed a large displacement precision XY positioning stage by
using the cross strip flexure joints and over-constrained mechanism.
According to previous work [13–18], the motion accuracy of such mechanisms is generally about 1% of its workspace. In traditional
design, there is no strict compensation for the coupled displacement among axes. So, such structures exhibit a large amount of cou-
pling displacement. For example, Lin and Lee [18] put forward a comprehensive flexible structure of a large-stroke compliant mech-
anism, which is a flexible mechanism based on pseudo-symmetrical structure. Simulation results show that its workspace is 2.34 mm.
Mechanism and Machine Theory 95 (2016) 125–139
⁎ Corresponding author. Tel.: +853 8822 4278; fax: +853 8822 2426.
E-mail address: qsxu@umac.mo (Q. Xu).
http://dx.doi.org/10.1016/j.mechmachtheory.2015.09.003
0094-114X/© 2015 Elsevier Ltd. All rights reserved.
Contents lists available at ScienceDirect
Mechanism and Machine Theory
journal homepage: www.elsevier.com/locate/mechmt

However, the amount of coupling displacement in non-working direction is almost 390 μm. That is, the cross-axis coupling amounts to
16.6%. Therefore, to achieve a precise positioning, it is necessary to reduce the crosstalk value.
On the other hand, it is challenging to devise a compliant micropositioning stage with a stroke longer than 10 mm in each working
axis. One reason lies in that the size of the flexible component is the decisive factor which determines its workspace directly [19]. It is
not easy to devise an XY stage having both large workspace and compact dimension simultaneously [19]. In recent research, a compact
stage is designed in [20], which has a dimension of 100 × 100 mm2
. Yet, the stage can just provide a workspace of 2 × 2 mm2
. In
contrast, the XY stage designed in [21] produces a workspace of 20 × 20 mm2
, but it is contributed by a large physical dimension
of 432 × 432 mm2
.
To this end, the conceptual design of a new decoupled XY micropositioning stage is proposed in this research. It can deliver a large
stroke and possesses a relatively compact structure thanks to a two-layer design. Specifically, the Roberts mechanism is introduced to
design a linear guiding mechanism. Considering that a single Roberts mechanism can only provide the linear motion of a single point,
two Roberts mechanisms working in parallel is employed to construct a flexible linear guiding mechanism. By cascading two newly
proposed linear compliant guiding mechanisms, a symmetrical structure design yields a novel large-stroke compliant parallel-
kinematic XY micropositioning stage based on Roberts mechanism. To the knowledge of the authors, no work has been conducted
to devise a multi-axis micropositioning stage using the Roberts mechanism in the literature.
The main contribution of this paper is the conceptual design and verification of a large-stroke parallel-kinematic compliant
XY micropositioning stage based on Roberts mechanism, which has not been attempted in previous literature. The remaining
parts of the paper are organized as follows. Section 2 presents the conceptual design of the new XY stage. Analytical modeling of
the XY stage is conducted in Section 3. These models are verified by conducting finite-element analysis (FEA) simulations in
Section 4. Section 5 describes a fabricated prototype and experimental investigations. Concluding remarks are summarized in
Section 6.
2. Conceptual design of large-range XY compliant mechanism
In this section, the traditional Roberts mechanism is introduced. As a traditional rigid-body mechanism, it can convert rotational
motion into straight-line motion approximately. Based on this mechanism, two novel compliant parallel linear guiding mechanisms
are proposed.
2.1. Background of Roberts mechanism
Roberts mechanism was first proposed by Richard Roberts (1789–1864) [22]. Its schematic is shown in Fig. 1(a). When the link
lengths are designed as R1 = R3 and the central blue triangle is isosceles, i.e., the two non R2 sides are of equal length, the trajectory
of point P will become an approximate straight line, as shown in Fig. 1(b).
2.2. Proposal of new compliant linear guiding mechanism
By replacing the conventional kinematic pairs of traditional rigid-body Roberts mechanism with flexible leaf springs, a new kind of
compliant linear guiding mechanism is obtained, as shown in Fig. 2(a). However, a single Roberts mechanism can only provide the
linear motion of a single point. To overcome this issue, two Roberts mechanisms connected in parallel can be employed to generate
Fig. 1. Schematic diagram of the Roberts Mechanism. (a) Traditional Roberts mechanism; (b) Structure of the Roberts mechanism.
126 S. Wan, Q. Xu / Mechanism and Machine Theory 95 (2016) 125–139

a compliant parallel linear guiding mechanism. According to the joining way, two different mechanisms are obtained as shown in
Fig. 2(b).
2.3. Design of new compliant serial XY linear guiding mechanism
Based on the two new flexible linear guiding mechanisms [see Fig. 2(b)], a novel XY mechanism is designed as shown in Fig. 3(a). It
is observed that the two compliant Roberts mechanisms are connected in series. In addition, the proposed design is a two-layer struc-
ture, and it can yield decoupling motion in the X and Y axes. The two-layer design enables the generation of a relatively compact di-
mension design.
In order to reduce the weight and to improve the resonant frequency, a lightweight optimization process is applied to the serial XY
guiding mechanism, as illustrated in Fig. 3(b). The resulted serial XY mechanism is employed in the following design of an XY sage.
2.4. Design of new compliant parallel-kinematic XY micropositioning stage
As a conceptual design, four serial XY linear guiding mechanisms are used to construct a parallel XY micropositioning stage, which
owns a 4-PP (P stands for prismatic joint) parallel mechanism structure. The CAD model of the overall structure for an XY stage is
shown in Fig. 4. This structure meets the requirement of integrated machining and easy assembly. In addition, the lightweight opti-
mization process reduces the mass of the platform and improves the dynamic performance of the system. Moreover, the idea of sym-
metrical structure is also adopted in the stage design. Such a symmetrical design can compensate for the coupling displacement to
some extent, as revealed in the discussion later. A specific dimension design is introduced in the following section by establishing
theoretical models.
(a) (b)
Fig. 2. (a) Flexure Roberts mechanism; (b) two new compliant Roberts linear guiding mechanisms.
Fig. 3. A novel serial XY flexure mechanism. (a) XY mechanism; (b) lightweight optimization process.
127
S. Wan, Q. Xu / Mechanism and Machine Theory 95 (2016) 125–139

3. Analytical modeling of the XY stage
3.1. Pseudo-rigid-body model (PRBM) of the compliant Roberts mechanism
As discussed above, the parallel-kinematic XY compliant stage consists of four serial XY compliant Roberts mechanisms. For each
single compliant Roberts mechanism, the kinematic principle is illustrated in Fig. 5. When an external force is applied to the coupler
point in the extension direction, the two ﬂexure beams are deformed to generate an approximate straight-line motion at the coupler
point.
To calculate the stiffness of a Roberts mechanism, the pseudo-rigid-body model (PRBM) can be used. The concept of PRBM is to
model the force-displacement relationship of a compliant element using an equivalent rigid-body mechanism along with the compli-
ance modeled as torsional springs around the joints. This allows the vast knowledge of traditional mechanism kinematics to be ap-
plied to the design of a compliant element.
Fig. 4. A new compliant parallel XY large-stroke micropositioning stage. (a) Top view; (b) isotropic view; (c) side view.
Fig. 5. Kinematic principle of a ﬂexible Roberts mechanism.

The PRBM of a single beam is shown in Fig. 6(a). In addition, the PRBM of a single flexure Roberts mechanism is shown in Fig. 6(b),
where the two flexure beams are replaced by two rigid links with four identical torsional springs [1]. The overall stiffness for this group
of flexure joints can be obtained through
k ≈
F
Δx
ð1Þ
where Δx and F are the displacement and applied force for the movable plate, respectively. These two parameters are obtained by the
following Eqs. (2) and (3).
Δx ¼ r2 cos θ2− cos θ20
ð Þ þ
r3
2
cos θ3−1
ð Þ þ b3 sin θ3 ð2Þ
F ¼ 4Kθ
2−h32
ð ÞΔθ2 þ 2h32
ð ÞΔθ4−
1 þ h42−h32
ð ÞΔθ3

r2 sin θ2 þ h32
r3
2
sin θ3−b3 cos θ3

h i−1
ð3Þ
Where b3 is the perpendicular distance from the center of link 3 to the coupler point [see Fig. 6(b)]. In addition, ri, θi, and θi0
(i = 2, 3, 4) are the pseudo-rigid-body (PRB) length, final angle, and initial angle of link i, respectively. The values for r2 and r3
are defined as γl, as shown in Fig. 6(b). Additionally, Kθ, h32, and KΘ are the kinematic coefficients which are defined by the fol-
lowing equations [23].
h32 ¼
δθ3
δθ4
¼
ω3
ω2
¼
r2 sin θ4−θ2
ð Þ
r3 sin θ3−θ4
ð Þ
ð4Þ
h42 ¼
δθ4
δθ2
¼
ω4
ω2
¼
r2 sin θ3−θ2
ð Þ
r4 sin θ3−θ4
ð Þ
ð5Þ
Fig. 6. The PRBM of compliant Roberts mechanism. (a) The PRBM of a single beam; (b) the PRBM of a single Roberts mechanism; (c) RBM for the Roberts mechanism.
129

Kθ ¼ γKΘ
Ech
3
6l
ð6Þ
Where γ is the characteristic radius factor, and KΘ is the stiffness coefficient. For this PRBM approximation, the values of γ and KΘ are
set as 0.852 and 2.65, respectively [25]. The terms Δθ2, Δθ3, and Δθ4 are the differences of link angles with respect to their initial positions.
In order to determine the required force for a given displacement, the instantaneous angles of the three links should be known. Given
an input displacement, all the angles can be found by solving (2) and the explicit four-bar equations simultaneously [1]. Then, these
angles can be substituted into (3) to find the required force. The resulting stiffness K of the flexure joint is determined by (1).
For an XY stage design as shown in Fig. 4(a), a stiffness model is derived as shown in Fig. 7. Then, the stiffness of the XY stage seen
at the input end can be computed as:
Kin ¼ 8k: ð7Þ
So, the driving force Fmax in one-sided direction can be calculated as follow.
Fmax ≥8k Dmax ð8Þ
Based on the pseudo-rigid-body model, the above mathematical models are developed relying on the hypothesis of small defor-
mation. Hence, they are acceptable for relatively small displacements only. If large deformation is considered, the elliptic integral ap-
proach can be employed to establish complete models [1].
3.2. Dynamics modeling
For a single compliant Roberts mechanism, four torsional springs (Kd1 = Kd2 = Kd3 = Kd4) and two lumped masses of the two flex-
ible links (lm2 = lm4) are added in the PRBM, as shown in Fig. 6(c).
Considering that m2 = m4, J2 = J4, r2 = r4, and lm2 can be calculated as:
Ji ¼
1
12
mil
2
i ð9Þ
lm2 ¼ lm4 ¼
1
4
m2 þ
J2
r2
2
ð10Þ
where m2 = m4 = l × c × h × ρAL with ρAL denoting the density of the material.
Fig. 7. Stiffness model of the XY stage.

The kinetic energy for the PRBM system can be expressed as follows [24]:
T ¼
1
2
meBx
2
¼
1
2
lm2vB
2
þ
1
2
lm4vc
2
þ
1
2
m3vm3
2
þ
1
2
J3ω3
2

þ
1
2
mb3vb3
2
þ
1
2
Jb3ω3
2

ð11Þ
where meB is the equivalent mass of a single compliant Roberts mechanism, mi and Ji are the mass and moment of inertia of link i
(i = 3, b), respectively. lm2 and lm4 are the lumped masses of the two flexible links, which are given in [26]. In addition, ẋ is the ve-
locity of the working platform in one working direction (X-axis), ω3 is the angular velocity of link 3, and vB, vC, vm3 and vb3 are the
velocities of the corresponding points, respectively.
From the kinematic analysis of the rigid-body motion of a four-bar linkage, it is known that
vB ¼ vC ¼ f θ2; x

; vm3 ¼ g θ2; x

; vb3 ¼ h θ2; x

; ω3 ¼ q θ2; x

ð12Þ
In addition, the elastic potential energy of the system in the X-axis working direction can be written as:
V ¼
1
2
Kinx
2
ð13Þ
The dynamics equation is generated by substituting the expressions of the kinetic and potential energies into the Lagrange's
equation:
d
dt
∂T
∂ x

−
∂T
∂x
þ
∂V
∂x
¼ Fx ð14Þ
where Fx is the generalized force along the X-axis of the system.
In order to calculate the natural frequency of the system, Fx = 0 is set in (14). Then, the free motion of the XY stage in one working
direction (X-axis) is governed by the following dynamic equation:
8meB þ mmov
ð Þ €
x t
ð Þ þ Kinx t
ð Þ ¼ 0 ð15Þ
where mmov is the equivalent mass of a single compliant Roberts mechanism, which is expressed by [22]:
meB ¼ α lm2 þ lm4
ð Þ þ βm3 þ ψmb3 þ
1
χ2
J3 þ Jb3
ð Þ ð16Þ
with χ ¼ r3
2 tg θ2 þ b3, α ¼ r3
2
4χ2
cos
2
θ2
, β ¼ ð1− b3
χ Þ
2
, and ψ ¼ ð1− b3
2χÞ
2
.
Based on the theory of vibration, the natural frequency (in unit of Hertz) of the XY stage in the X-axis working direction is com-
puted as follows.
f x ¼
1
2π
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Kin
8meB þ mmov
s
ð17Þ
Due to the symmetry of the parallel-kinematics XY stage, the natural frequency fy of the stage in the Y-axis direction is equal to fx. In
order to generate a high bandwidth of the control system, a high resonant frequency of the stage is desired.
4. Model verification with FEA simulations
The XY stage is driven by two voice coil motors (VCMs) to produce the translation in XY plane. Given a driving force, the position of
the XY stage is determined by the link dimensions (which determine the stiffness) of the mechanism and the external forces applied
to the mechanism. In this work, the XY stage is designed to generate a stroke over ±7 mm in each working direction as an illustration.
For the convenience of calculation, the angle parameters are selected as θ20 = θ40 = 75∘
, r3 = 30 mm, and θ30 = 180∘
. Moreover, the
natural frequency of the system is constrained to be greater than 40 Hz. Such constraint is imposed by using the dynamic model (17).
Based on the design criteria (2), (3), (8), and (17), the models described above can be used to determine the unknown geometric
parameters of the structure. As a case study, the chosen geometry parameters, which satisfy the design criteria, are summarized in
Table 1.
4.1. Static performance analysis with FEA
The quantitative models are developed under the ideal situations. In order to test the performance of the designed stage and to
verify the accuracy of the derived models, FEA simulation is conducted by using ANSYS software package. The material of the stage
131

is selected as Al-7075 alloy, which is easy to machine, and its main characteristic parameters are: Young's modulus = 71.7GPa, yield
strength = 503 MPa, Poisson's ratio = 0.33, and density = 2.81 × 103
kg/m3
.
The quantitative models as developed in previous section predict that, the work range of the XY stage in each direction is ±7 mm.
The deformation result is shown in Fig. 8(b). In addition, Eq. (8) evaluates that the required maximum force to drive the stage is
Fmax = 180.5 N. However, the FEA simulation result shows that the required actuation force is 160 N. Taking the FEA result as the
Fig. 8. Static FEA results of the XY stage under different conditions. (a) Deformation result when δy = 7 mm and δx = 0 mm; (b) Deformation result when δy = 7 mm
and δx = 7 mm.
Table 1
Geometry parameters for the mechanism.
Parameter (Symbol) Value
Length of flexible links 2 and 4 (l2, l4) 36 mm
Width of flexible links 2 and 4 (c2, c4) 10 mm
Thickness of flexible links 2 and 4 (h2, h4) 0.5 mm
Length of flexible link 3 (l3) 30 mm
Width of flexible link 3 (c3) 10 mm
Thickness of flexible link 3 (h3) 3 mm
Length of flexible link b3 (lb3) 36 mm
Width of flexible link b3 (cb3) 10 mm
Thickness of flexible link b3 (cb3) 6 mm
Initial angle of links 2 and 4 75°

benchmark, it is seen that the established quantitative models predict the stage performance with a deviation of 12.8% with respect to
FEA result.
Moreover, when one input displacement arrives at 7 mm, the deformation result of FEA static simulation is illustrated in Fig. 8(a). It
shows the compliance of the Y-axis direction when δx = 0. The yield strength of the Al-7075 is [σ] = 503 MPa. As can be seen from
Fig. 9, the induced maximum stress is 485.25 MPa b [σ] = 503 MPa, which shows a safety factor of 1.27 of the material, meeting the
strength condition.
In addition, Fig. 10 shows the compliance in the Y-axis direction when δx = 0. As compared with the FEA result, the result of the
developed quantitative model is 5.6% lower. The discrepancy mainly comes from the small deformation assumptions of the PRBM.
On the other hand, the FEA results as illustrated in Fig. 11 shows the parasitic motion in Y-axis, which is induced by the actuation in
X-axis. Fig. 11(a) exhibits the FEA results of the symmetrical (4-PP) and asymmetric (2-PP) structure designs. It is obvious to observe
that as compared with the asymmetric design, the symmetrical structure design can significantly reduce the parasitic motion in the
non-working direction by over 60 times.
The new Roberts mechanisms can produce a straight-line motion under the working situation. As shown in Fig. 11 and Table 2,
when the stage delivers the full work range in the working direction (i.e., 7 mm in X-Axis), the parasitic motion of the central
point in the non-working direction of the output stage (i.e., Y-Axis) is 0.64 μm. It means that the maximum cross-axis displacement
caused by the work direction is only 0.09% in theory.
4.2. Dynamic performance analysis with FEA
The FEA results of dynamic performance analysis are shown in Fig. 12. It is observed that the first-three natural frequencies are
32.3 Hz, 32.5 Hz, and 58.7 Hz, respectively. As described in the previous section, the designed natural frequency is 40 Hz. This reveals
that the developed quantitative dynamic model predicts the stage's natural frequency with a deviation of 18.75% with respect to FEA
result.
Besides, the modal shapes of the first two modes are shown in Fig. 12(a) and (b), which are the translations along the two working
axes. The third one as shown in Fig. 12(c) is the rotation of the output platform in the working plane. The first-six resonant frequencies
are shown in Table 3. The similar values of the first-two natural frequencies indicate that the XY stage has almost the same dynamics
performance in the two working directions, which benefits from the parallel-kinematic design of the stage. In addition, the frequency
of the rotational mode is almost twice higher than the first-two modes, which confirms that the stage has two desired degrees of
freedom.
5. Prototype fabrication and experimental study
5.1. Prototype fabrication
The assembled CAD model of the designed XY stage is shown in Fig. 13(a). The XY stage has been fabricated using Al-7075 alloy
through the wire electrical discharge machining (WEDM) process. The prototype is shown in Fig. 13(b), which is driven by two
VCMs. The output positions are measured by two Keyence laser displacement sensors. The controller is constructed by NI CRIO
9022 and 9018 FPGA module. Control algorithm is realized by NI LABVIEW software package.
Fig. 9. FEA results of stress distribution.
133

Fig. 11. Comparison of parasitic motions for the XY stage. (a) FEA simulation results of parasitic motions of the XY Stage with and without symmetrical structure design;
(b) experimental results of parasitic motions of the XY stage with symmetrical structure design.
Fig. 10. The motion displacement comparison between FEA and calculated results.

Fig. 12. FEA simulation results of modal analysis. (a) 1st
resonant mode; (b) 2nd
resonant mode; (c) 3rd
resonant mode.
Table 2
FEA simulation results of the parasitic motion with and without symmetrical structure design.
Working displacement
in X-axis (mm)
The parasitic motion in Y-axis without
symmetrical structure design (μm)
The parasitic motion in Y-axis with
symmetrical structure design (μm)
0 0 0
1 6.0361 0.0839
2 12.091 0.1551
3 17.656 0.2297
4 22.879 0.3552
5 29.362 0.4517
6 35.529 0.5429
7 41.076 0.6371
135

Since the FEA result reveals that Fmax = 160N, two VCM motors (model: NCC05-18-060-2X, from H2W Techniques, Inc.) are se-
lected to drive the stage. Each VCM can provide a sufficient large output force of 194.6 N with a stroke of 12.7 mm. The total dimension
of the stage is 244 × 244 mm2
. So, the area ratio (i.e., the ratio between the planar workspace and physical dimension) of the stage is
0.32%. It is seen that the proposed stage exhibits a relatively large area ratio, which means that the space is made full use by the two-
layer design.
The two-layer design enables the generation of relatively compact physical size in comparison with a monolithic design. However,
the shortcoming of two-layer design lies in that assembly is needed to develop the XY stage. Considering that the fixing holes can be
manufactured with a high precision of ±5 μm, a low level of assembly error is expected. The performance of the developed XY stage is
tested through the subsequent experimental studies.
5.2. Experimental investigations
In this section, experimental studies are carried out to test the statics and dynamics performance of the developed XY
micropositioning stage.
First, the stroke of the XY stage is tested by applying a 0.1-Hz, 7-V sinusoidal voltage signal to drive the each VCM at a time. For
instance, by applying the signal as shown in Fig. 14(a) to the VCM in X-axis, the output displacements in X and Y axes are shown
in Fig. 14(b) and (c), respectively.
It is seen from Fig. 14 that the motion range in X-axis is 12.4 mm, and the caused coupling motion amounts to 195.0 μm.
This indicates a crosstalk of 195μm
12:4mm 100% ¼ 1:57% between the axes. Similar results are produced when the VCM in Y-axis is
actuated. That is, the motion range in Y-axis is 12.2 mm, and the coupling motion in X-axis is 202.3 μm, which exhibits a crosstalk
of 202:3μm
12:2mm 100% ¼ 1:66% between the two working axes.
It is observed that the experimental results (1.57% and 1.66%) of crosstalk is much bigger than the FEA simulation result (0.09%).
These discrepancies are mainly caused by the manufacturing tolerance and assembly error of the XY stage. Even so, the assembled XY
stage is capable of providing a smooth motion without backlash and clearance, which is enabled by the compliant mechanism design.
Hence, the device can be used for micropositioning application. If the machining and assembly precision can be improved, the discrep-
ancy between the simulation result and experiment result will be reduced. In practice, to reduce the crosstalk values and to achieve
precise positioning, suitable feedback control algorithm can be implemented to cooperatively control the output displacements in the
two working axes for the XY micropositioning stage.
Next, the dynamics performance of the XY stage is examined by applying a swept-sine voltage signal to drive each VCM alternately.
By driving the X-axis VCM using a swept-sine voltage with the frequency ranging from 0.1 to 1000 Hz, the produced X-axis motion
response is recorded by the laser displacement sensor. The frequency response is obtained by FFT algorithm and the result is
Fig. 13. CAD model and fabricated prototype of the XY micropositioning stage. (a) CAD model; (b) developed prototype.
Table 3
The first six frequencies of resonant modes.
Mode Frequency (Hz)
1 32.332
2 32.527
3 58.758
4 212.74
5 324.33
6 326.42

shown in Fig. 15. It is found that the stage exhibits a natural frequency of 18.6 Hz in the X-axis. Moreover, the natural frequency in the
Y-axis is obtained as 19.5 Hz, which is similar to the X-axis value.
5.3. Discussion on stage performance
The experimental results demonstrate a large stroke of the developed XY micropositioning stage. The crosstalk between the two
working axes may be induced by the fabrication errors of the stage parameters, and assembly errors of the stage and motors. In
0 1 2 3 4 5 6 7 8 9
-10
-5
0
5
10
Time (s)
Input
voltage
(V)
(a)
0 1 2 3 4 5 6 7 8 9
-1
-0.5
0
0.5
1
x 10
4
Time (s)
Displacement
x
(µm)
(b)
0 1 2 3 4 5 6 7 8 9
-200
-100
0
100
Time (s)
Displacement
y
(µm)
(c)
Fig. 14. Experimental results of X-axis motion range testing. (a) Voltage applied to VCM; (b) output displacement in X-axis; (c) output displacement in Y-axis.
10
-1
10
0
10
1
10
2
10
3
-50
0
50
100
Magnitude
(dB)
10
-1
10
0
10
1
10
2
10
3
-2000
-1000
0
1000
2000
Phase
(degrees)
Frequency (Hz)
Fig. 15. Frequency response of the XY stage in X-axis direction.
137

addition, due to the stroke limit of the adopted VCM motors, the experimental result of the XY stage stroke is slightly lower than the
analytical result. The desired range of ±7 mm can be achieved by employing VCMs with larger stroke.
The experimental results reveal that the XY stage exhibits the natural frequencies of 18.6 and 19.5 Hz in X- and Y-axes, respectively.
Such values are much smaller than the FEA results of 32.3 and 32.5 Hz in the two working axes, respectively. These discrepancies are
mainly caused by (1) the mass of the moving coil of VCM motor, which is not considered in FEA simulation, and (2) the fabrication
errors of the flexure hinges. For illustration, by considering the VCM motors in FEA, the simulation results of the first two resonant
modes are shown in Fig. 16. It is found that the first two resonant frequencies are 26.04 and 20.06 Hz, which are closer to the exper-
imental results. As compared with these simulation results, the experimental results of the resonant frequencies are about 25% lower.
These deviations are mainly induced by the fabrication errors of the XY stage.
In the future, a higher natural frequency will be achieved by implementing an optimal design of the XY stage parameters. In addi-
tion, control schemes will be implemented to achieve a precise positioning control of the XY stage. As a new attempt, it is notable that
the designed XY stage belongs to plane mechanisms and it only provides in-plane translational motion. The two-layer design is
employed to generate a compact physical dimension of the XY stage. In the future, an XYZ stage will be designed to provide three-
dimensional translations in space. In addition, as compared with existing parallelogram linear motion compliant mechanisms, it
has been shown that the compliant Roberts mechanism can be employed to construct a variable-stiffness mechanism by varying
the effective length of the compliant links [27]. In the future, an XY stage with variable stiffness will be designed for relevant
applications.
6. Conclusion
A new type of parallel-kinematic XY micropositioning stage based on flexure Roberts mechanism has been designed, analyzed,
fabricated, and tested in this paper. FEA simulation results show that this kind of compliant mechanism not only has good perfor-
mance in reducing the parasitic motion in non-working direction, but also can produce a large workspace of 14 × 14 mm2
. So, it
meets the requirements of a large stroke, high precision linear guiding mechanism. The large stroke and small crosstalk have been
Fig. 16. FEA simulation results of modal analysis for the XY stage with actuators. (a) 1st
resonant mode; (b) 2nd
resonant mode.

veriﬁed by experimental studies. Experimental testing demonstrates a natural frequency around 20 Hz in both working axes.
Moreover, it owns a relatively compact structure thanks to a two-layer mechanism design. It is found that it is feasible to devise a
large-stroke micropostioning stage based on the Roberts mechanism. Future work involves architectural optimization and precision
control of the XY stage for pertinent applications.
Acknowledgments
The work is support by the National Natural Science Foundation of China under Grant No. 51575545, the Macao Science and
Technology Development Fund under Grant No. 052/2014/A1, and the Research Committee of the University of Macau under Grant
No. MYRG078(Y1-L2)-FST13-XQS.
References
[1] L.L. Howell, Compliant mechanisms, Wiley, New York, 2001.
[2] Y.K. Yong, T.F. Lu, Relationship among input-force, payload, stiffness and displacement of a 3-DOF perpendicular parallel micro-manipulator, Mech. Mach. Theory
45 (2010) 756–771.
[3] Q. Xu, Design and development of a flexure-based dual-stage nanopositioning system with minimum interference behavior, IEEE Trans. Autom. Sci. Eng. 9 (3)
(2012) 554–563.
[4] H. Wang, X.M. Zhang, Input coupling analysis and optimal design of a 3-DOF compliant micro-positioning stage, Mech. Mach. Theory 43 (2008) 400–401.
[5] Q. Xu, Design and testing of a novel multi-stroke micropositioning system with variable resolutions, Rev. Sci. Instrum. 85 (2) (2014) 025002.
[6] S. Polit, J. Dong, Development of a high-bandwidth XY nanopositioning stage for high-rate micro/nano manufacturing, IEEE/ASME Trans. Mechatron. 16 (4)
(2011) 722–724.
[7] K.-B. Choi, H.J. Lim, G.H. Kim, J.J. Lee, A flexure-based scanner for a fully bidirectional operation driven by a differential piezo force, Proc. IMechE C J. Mech. Eng. Sci.
(2014)http://dx.doi.org/10.1177/0954406214525602.
[8] K. Tan, G.H. Zong, S.S. Bi, Z. Yu, Multi-leaf configuration of expanded-motion-range flexure hinges, Dual Use Technol. Prod. 3 (2007) 38–39.
[9] S.S. Zhao, S.S. Bi, G.H. Zong, J. Yu, New large-deflection flexure pivot based on curved flexure element, Chin. J. Mech. Eng. 45 (4) (2009) 8–12.
[10] Q. Xu, Design, testing and precision control of a novel long-stroke flexure micropositioning system, Mech. Mach. Theory 70 (2013) 209–224.
[11] B.P. Trease, Y.M. Moon, S. Kota, Design of large displacement compliant joints, J. Mech. Des. 127 (4) (2004) 788–798.
[12] Y.J. Choi, S.V. Sreenivasan, B.J. Choi, Kinematic design of large displacement precision XY positioning stage by using cross strip flexure joints and over-constrained
mechanism, Mech. Mach. Theory 43 (6) (2008) 724–737.
[13] X.D. Shi, J.L. Li, L.W. Wang, Design of large displacement and high precision linear stage based on laminated flexure hinge and ultrasonic motor, Manuf. Technol.
Mach. Tool 2 (2007) 104–106.
[14] S. Awtar, A.H. Slocum, Constraint-based design of parallel kinematic XY flexure mechanisms, J. Mech. Des. 129 (8) (2007) 816–830.
[15] D. Devaprakasam, S.K. Biswasa, Design of a flexure for surface forces apparatus, Rev. Sci. Instrum. 74 (3) (2003) 1228–1235.
[16] X. Tang, I.-M. Chen, Q. Li, Design and nonlinear modeling of a large-displacement XYZ flexure parallel mechanism with decoupled kinematic structure, Rev. Sci.
Instrum. 77 (11) (2006) 115101.
[17] Q. Xu, Design and development of a compact flexure-based XY precision positioning system with centimeter range, IEEE Trans. Ind. Electron. 61 (2) (2014)
893–903.
[18] Y.T. Lin, J.J. Lee, Structural synthesis of compliant translational mechanisms, Proc. 12th IFToMM World Congress 2007, pp. 1–5.
[19] Q. Xu, New flexure parallel-kinematic micro-positioning system with large workspace, IEEE Trans. Robot. 28 (2) (2012) 478–491.
[20] D. Kang, K. Kim, D. Kim, J. Shim, D.G. Gweon, J. Jeong, Optimal design of high precision XY-scanner with nanometer-level resolution and millimeter-level working
range, Mechatronics 19 (4) (2009) 562–570.
[21] Z. Li, J. Yu, A novel large-range XY compliant parallel micro-manipulator, J. Chang. Univ. 25 (4) (2013) 14–19.
[22] A.G. Ambekar, Mechanism and machine theory, Prentice-Hall of India, New Deli, 2007.
[23] W. Chen, J. Jiang, W. Chen, J. Liu, A novel flexure-based uniaxial force sensor with large range and high resolution, Sci. China Technol. Sci. 58 (8) (2013)
1940–1948.
[24] W. Chen, J. Jiang, J. Liu, S. Bai, W. Chen, A passive eddy current damper for vibration suppression of a force sensor, J. Phys. D. Appl. Phys. 46 (2013) 075001.
[25] N.B. Hubbard, J.W. Wittwer, J.A. Kennedy, D.L. Wilcox, L.L. Howell, A novel fully compliant planar linear-motion mechanism, Prof. ASME 2004 International Design
Engineering Technical Conferences and Computers and Information in Engineering Conference, Salt Lake City, Utah, USA 2004, pp. 1–5.
[26] Y.Q. Yu, L.L. Howell, C. Lusk, Y. Yue, M.G. He, Dynamic modeling of compliant mechanisms based on the pseudo-rigid-body model, J. Mech. Des. 127 (2005)
760–765.
[27] J.C. Hawks, M.B. Colton, L.L. Howell, A variable-stiffness straight-line compliant mechanism, Prof. ASME 2015 International Design Engineering Technical Conferences
and Computers and Information in Engineering Conference, Boston, Massachusetts, USA 2015, pp. 1–10.
139

wan2016548gfdruoohfrewwwryuopjbvfddss.pdf

More Related Content

Similar to wan2016548gfdruoohfrewwwryuopjbvfddss.pdf (17)

Recently uploaded (20)

wan2016548gfdruoohfrewwwryuopjbvfddss.pdf