Embed Size (px)
Citation preview
electro
Abstract— Tactile display has recently been attracting much
attention in the field of human computer interaction. There is a
strong need for such a device especially for surgeons who need
to feel the tissue hardness during laparoscopic surgeries. The
concept of design is built upon simple use of several springs
which are made from Shape Memory Alloy, SMA, to control
shape and stiffness. In this paper, we describe the design of a
new device which can display both surface shape and stiffness
of an object. The design parameters of this device are selected
based on the spatial resolution of human finger and the
stiffness range of the soft tissue. The design concept was
validated through finite element analysis for the selected
parameters.
I. INTRODUCTION
ACTILE displaying of real objects in virtual
environments has recently attracted considerable
research interest in the rapidly developing field of virtual
reality technology [1]. A tactile display device is a human–
computer interface that can reproduce, as closely as possible,
the tactile parameters of an object, such as shape, softness,
surface texture, roughness, vibration and temperature [2].
For humans, tactile sensations are perceived through
mechanoreceptive units embedded in the outer layers of the
skin. Which in turn transmit signals to the brain when
stimulated. Tactile displays can serve as an alternative for
information transmission through the stimulation of the
human skin to induce tactile perception [3], [4].
The literature of tactile display devices is huge and
diverse. The basic concept used depends on different
techniques including electromagnetic, piezoelectric,
magneto rheological fluids, electro rheological fluids, shape
memory alloy, and conjugated polymers techniques. Devices
using electromagnetic fields are affordable and easy to be
controlled. The highest possible delivered forces were
around 1.7N [5], [6]. Tactile display systems based on
piezoelectricity give low displacements with high bandwidth
and need high voltage [7]. Depending on this property,
research was done in (Stimulator for Tactile Receptors by
The first author had been supported by Mitsubishi Corporation
International scholarship which is gratefully acknowledged.
Nader A. Mansour, M.Sc. student at Egypt-Japan University for Science
and Technology, Mechatronics and Robotics dept., New Borg-El-Arab city,
Alexandria, Egypt (phone: 0020100-990-5122; Zip: 21934; e-mail:
Ahmed M.R. Fath El-Bab, assistant professor at Faculty of Engineering,
Assiut University, Mechanical Engineering Dept., Egypt (e-mail:
Mohamed Abdellatif, associate professor at Egypt-Japan University of
Science and Technology, Mechatronics and Robotics dept., New Borg-El-
Arab city, Alexandria, Egypt (e-mail: [email protected]).
Skin Stretch) STRess [8] using bimorphs to stretch the skin
laterally. It could deliver only 25 micrometer but in recent
work [9], it is reported to be more robust and reached
0.1(mm) deflection. It can also deliver different bandwidths
in the range 0:1 (kHz).
Active fluids also could be used in tactile display systems
like the Electro Rheological Fluids (ERF) and the Magneto
Rheological Fluids (MRF), in which the viscosity of the
fluid changes when an electric or a magnetic field is applied.
High voltage is required (maximum 10 kV per 20 mA) and
fabric is needed between the layers to prevent short circuit
and to increase the force output [10].
Concerning the conjugated polymers, they consist of
electrolyte between conductive polymers. They have a
powerful actuator with up to 15% strain and 49(MPa) and a
stretchable actuator with 34% strain and up to 10(MPa), but
not yet tried [11], [12].
Another approach used the Shape Memory Alloy (SMA) as
an actuator for tactile display systems. SMA's are used for
their high power to volume, power to weight and force to
weight ratios.
According to the aforementioned work, displaying only one
property of the following could be implemented;
1- Display the shape of an object.
2- Display mechanical vibration for tactile feedback.
3- Display of softness of an object (very limited trial
using ERF and MRF).
In this work, the conceptual design of a multimodal tactile
display device for displaying both the shape and stiffness of
an object are presented. The display device consists of two
springs, Elongation Spring (ES) and Stiffness Spring (SS)
for displaying shape and stiffness, respectively. The spring
material is selected as SMA in order to control its elongation
and stiffness simultaneously. The spring parameters are
designed to display the stiffness of human/animal organs
(soft tissue). The device should be a part of an integrated
system in which the shape and softness are measured in the
object space and this information is transmitted to another
remote location, in which it is needed to reproduce this
information for the user space. Our work is focused only in
the user space while object space sensing is out of scope.
The paper is organized as follows; the next section
presents the proposed system description and principles of
operation. Section III describes the design considerations
and procedure. In section IV, the Finite Element Analysis
(FEA) is applied to validate the concept design. Conclusions
are then summarized in section V.
Design of a Novel Multi-Modal Tactile Display Device for
Biomedical Applications
Nader A. Mansour, Ahmed M.R. Fath El-Bab, and Mohamed Abdellatif
T
The Fourth IEEE RAS/EMBS International Conferenceon Biomedical Robotics and BiomechatronicsRoma, Italy. June 24-27, 2012
978-1-4577-1198-5/12/$26.00 ©2012 IEEE 183
II. CONCEPT OF THE MULTIMODAL TACTILE DISPLAY DEVICE
Fig. 1. describes the concept for the tele-transmission of
the tactile perception. In order to present shape and stiffness,
a matrix of a 5x5 pin units, with a spatial resolution of
5(mm) is arranged as shown in Fig. 2.
Each pin unit uses two decoupled SMA springs; one of
them is responsible for displaying the object shape/height
through changing its displacement (ES). The second spring
is responsible for displaying stiffness (SS). This stiffness
information can be transmitted only when the user presses
the display pin. These springs are located on series one
above the other, as shown in Fig. 3 So the human user can
feel the resultant change in both shape/height and stiffness
simultaneously.
The lower spring, ES,- squared and ground tension spring
depending on the Two Way Shape Memory Effect
(TWSME) that is trained to remember two different shapes;
the contracted shape in Martensitic phase and the elongated
shape in Austenitic phase so that it is responsible for
displaying the shape/height.
Fig. 1. The system for multi-modal tele-transmission of tactile perception.
Fig. 2. Schematic of the structure of Multimodal tactile display device.
Fig. 3. Tactile display pin unit and spring model.
Fig. 4. Tactile display of liver with abscess, (a) liver with abscess [13], (b)
tactile display of the liver shape and the abscess hardness.
The TWSME is selected here as it can elongate to display
shape when heated and it will return to its original shape
when cooled with no need for external bias spring for
driving it back.
This spring is settled on the medial plate and wound around
a collar that can move freely upwards and downwards inside
a guide in the medial plate.
The upper spring, SS, is squared and ground compression
spring depending on the One Ways Shape Memory Effect
(OWSME) that is trained to stiffen/soften while keeping its
original shape so that it is responsible for displaying only the
stiffness. This spring is settled on the collar and wound
around another pin of diameter 2(mm) that can move freely
upwards and downwards through the collar.
The collar can then be moved upwards and downwards by
applying electrical current to elongation spring ES that will
push the group of SS and the pin upwards and downwards
within a stroke of 10(mm). On the other hand, stiffness
spring SS can be actuated by applying electrical current to
control its stiffness. In this way, the tactile display system
can be configured to present the tactile information.
Upper base
Lower base
5x5 tactile
display pin
matrix
Collar ES
SS
Medial plate
Hard
Medium
Soft
(a)
(b)
Object Space
Signal Transmission
User Space
Tactile Display Tactile Sensor Abscess
Abscess
Liver
Collar
Stiffness
Spring
(SS)
Tactile display pin
Elongation
Spring
(ES)
184
The purpose of the device is to tactually display the shape
of human organs and display its softness/hardness in the
same time. Fig. 4 shows an example and application of the
device to display the shape of liver by changing the height of
each pin and to display the hardness of abscess in the soft
liver by increasing the stiffness of the middle pins.
The tactile display unit can be modeled as two springs in
series, as shown in Fig. 3. The resultant stiffness, kR, can be
computed as per the well known equation of:
(1)
where kst and kEl are the stiffness constants of the stiffness
and elongation springs, respectively. Subsequently, the
stiffness spring, SS, will be the dominant spring stiffness of
the whole system.
In the next section, we show how to evaluate the spring
stiffness to satisfy these conditions. kst and kEl can be chosen
so that the resultant stiffness kR is made equal to kst.
III. DESIGN CONSIDERATIONS
It should be noted that as for humans, the shape can be
sensed by touch, without exerting forces, or by vision sense
and the shape perception is usually separate from the
softness measurement mode. The switching to the stiffness
mode is usually done by exerting a reference force and
measuring the resulting displacement. It is obvious that
softness measurement mode alters the shape, which has already been measured. We want to implement this dual
mode measurement concept in our new device.
A. Stiffness display range
Stiffness display values depend on both, the display pin
dimension and the geometry of the tissue being displayed.
This concept is slightly different from the concept of the
Young’s modulus of a material which is unique and fixed
value for each material. Hayes et al. [14] presented a
mathematical model for the elasticity problem of indentation
test as follows:
(2)
where, F, w, a, ν and h, as shown in Fig. 5., are the applied
force (here this force will be applied by a finger), indentation
depth (tissue deflection), radius of the display pin, Poisson's
ratio of the tissue and the tissue thickness, respectively.
However, Ck is a scaling factor depends on ν, (a/h) and
(w/h).
By modifying (2), the stiffness of the soft tissue, which
will be presented by the tactile display device; (kst= F/w) can
be expressed as:
(3)
Poisson's ratio can be assumed to be 0.3 as a mean value.
Equation (3) can be used for determining the stiffness
measuring range based on the expected range of Young's
modulus and the scaling factor .
Young's modulus of 100 (kPa) is selected as the display
upper range based on the previously reported values by
several researchers. For instance, Chinzei et al. [15] showed
experimentally that the Young’s modulus of swine brain is
about 7.425 (kPa). Farshad et al. [16] performed a series of
aspiration experiments using a pig's kidney; the results
showed that the Young’s modulus of the pig's kidney was
43.5 (kPa). Kim et al. [17] developed a system for measuring
the mechanical properties of soft tissue of a pig in vivo. The
results showed that the Young’s modulus were 31.8 and 48.8
(kPa) for the liver and the esophagus, respectively.
Zheng and Mak [18], studied the biomechanical
properties of human lower limb soft tissues and reported that
the Young’s modulus ranges from 10.4 to 89.2 (kPa).
Fig. 5. Indentation model parameters.
In order to estimate the stiffness display range, it is
required to select the display pin diameter and the tissue
height. It is assumed that, the tissue height from a bony layer
to the outer skin layer should be less than 10 mm. The
display pin diameter is selected as 2 mm to suit the spatial
resolution of human fingertip (ranging 1.5 to 3 mm [19]).
Zhang et al. [20] estimated the value of Ck and reported it
to be 1.15 assuming ν = 0.3, a/h= 0.1, w/h = 50% through a
non linear (FEA). Using the values of pin diameter 2 (mm)
(a = 1mm), ν = 0.3 and Ck = 1.15, the maximum kst can be
obtained 250 (N/m) at E = 100 (kPa), and minimum stiffness
Kst of 85 (N/m). However, the value of the stiffness display
range can be changed according to the application.
B. Spring stiffness selection
The spring stiffness kst and kEl will be selected to display
stiffness of 250(N/m) as upper limit and one third of this
value; i.e. 85 (N/m) as lower limit. The selection of this ratio
is based on the available change ratio in the Young’s
modulus between the Martensite (E=28GPa) and Austenite
(E=80GPa) phases of SMA. According to this ratio, the
design will represent soft tissue elasticity ranged from 35 to
100(kPa), as calculated from (3).
As stated in section I, Elongation Spring, ES, will be
responsible for displaying the shape by elongating its length.
On other hand, Stiffness Spring, SS, will be responsible for
displaying the stiffness by changing its Young’s modulus
h Soft tissue
185
from 28 to 80 (GPa) during phase change from Martensite to
Austenite which is a unique privilege of the SMA material
over other materials. Table I summarizes the mechanical
properties of NiTi as a SMA material [21].
TABLE I: MECHANICAL PROPERTIES OF NITI
For displaying stiffness ranged from 85 to 250 (N/m) the
stiffness spring will be designed to give 85 (N/m) at
Martensite phase. And according to (1), kEl should be
selected so that the total stiffness KR equals kst. For this
purpose and assuming that kEl = N x kst, this will result in
KR= m x kst. By increasing N, the total stiffness becomes
almost equal to kst, as shown in Fig. 6. At N=10, kR= 0.9 kst
and kEl should be equal to 10 kst (N/m).
Fig. 6. The selection of ratio between (kEl/kst).
C. Springs dimensions selection
The dimensions of ES and SS will be selected to achieve the
above stiffness values. Spring stiffness, k (N/m) can be
modeled as per [24] by the following equation:
(4)
where: G is the spring modulus of rigidity. , and Na
are spring wire diameter, mean diameter, and number of
effective coils respectively, as shown in Fig.7.
Fig. 7. Schematic drawing of the spring parameters.
springs as the applied current to SMA is highly
dependent on so, the dimension of Elongation Spring,
ES, and Stiffness Spring, SS, could be determined using (2)
while keeping the spring index; C = / , within the
allowed range of 5 ≤ C ≤ 12 [22]. Table II shows the spring
dimensions and modulus of rigidity in the Martensite and
Austenite phases to achieve the desired stiffness/display
range.
TABLE II THE SPRINGS DIMENSIONS.
(GPa)
Na k
(N/m)
Stiffness
spring
( kst )
14
Martensite 0.3 2.6 10 85
40
Austenite 0.3 2.6 10 250
Elongation
spring
(kEl)
20
Martensite 0.8 4 7 2500
40
Austenite 0.8 4 7 4571
The parameters shown in table II are selected based on
10mm displacement range. For increasing this range the
number of turns; Na, should be increased and subsequently
the Dw, and Dm, should be increased for keeping the same
spring stiffness. This will increase the device size and
subsequently its spatial resolution will be lowered.
The designed spring parameters/dimensions will be
evaluated to decide its ability to display the desired stiffness
and shape ranges as in the next section.
IV. FINITE ELEMENT ANALYSIS
A finite element model was applied to simulate the
stiffness that will be displayed by the tactile display device.
This model was established as shown in Fig. 8. by meshing
the tissue into solid tetrahedral elements.
The material of springs ES and SS are assumed to be linearly
elastic with Poisson's ratio of 0.33. Large deformation was
taken into consideration in the analysis by choosing
nonlinear geometry and applying the load incrementally.
The load is applied as pressure on the pin top surface to
cause a force increased from 0 to 1N.
Martensite Austenite
Young’s modulus 20-45 (GPa) 30-83 (GPa)
Ultimate Tensile Strength 800-1900
(MPa)
800-1900
(MPa)
Elongation at Failure 20-25% 20-25%
Recoverable strain 8-10 % 8-10%
Poisson Ratio 0.33 0.33
0.2
0.4
0.6
0.8
1.0
Ratio between the stiffness of elongation and
stiffness spring N= (kEl/kst).
Ra
tio
betw
een
th
e r
esu
lta
nt
stif
fness
an
d s
tiff
ness
sp
rin
g m
= (
kR/k
st).
16 20 12 8 4
186
Fig. 8. Finite element Model.
Fig. 9. The stiffness change of spring SS while ES is kept not actuated.
Fig. 9. shows the simulation results of the force-
displacement relationship of the total system; kR, in two
situations;
1- The blue curve represents the system behavior while the
Young’s modulus of SS= 28(GPa) (to represent the
softness case in Martensite phase) and Young’s
modulus of ES= 40(GPa) (to represent the elongation
spring in the compressed situation, in Martensite phase),
as shown in Fig. 10(a).
2- The red curve represents the system behavior while the
Young’s modulus of SS= 80(GPa) (to represent the
Austenite phase, i.e. the hardness case) and Young’s
modulus of ES= 40(GPa) (to represent the elongation
spring in the compressed situation, i.e. Martensite
phase), as shown in Fig. 10 (b).
As shown in Fig. 9, the total stiffness of the system could
be changed from 95(N/m) to 275 (N/m) as a result of SS
actuation while keeping ES not actuated.
Another simulation was applied to check the system
behavior while actuating the ES. This second simulation
studied the stiffness change of SS in two situations;
1- Young’s modulus of SS = 28(GPa) (to represent the
softness case in Martensite phase) and Young’s
modulus of ES= 80(GPa) (to represent the maximal
elongation in Austenite phase), as shown in Fig. 10
(c) and it resulted in a curve almost the same as that
of the blue curve in fig. 9.
2- The Young’s modulus of SS= 80 (GPa) (to represent
the hardness case in Austenite phase) and Young’s
modulus of ES= 80(GPa) (to represent the elongation
spring in the elongated situation, i.e. Austenite phase),
as shown in Fig. 10 (d) and it resulted in a curve
almost the same as that of the red curve in fig. 9.
The simulation shows that total stiffness KR is the same as
that of fig. 9 with error in the stiffness not more than 1.2%.
According to these simulation results, it’s found that the
total stiffness KR is independent of the applied force on the display pin and independent of the elongation of the ES.
In the proposed device, the response depends on the rate
of phase transformation from martensite to austenite and
vice verse, which mainly depends on two factors;
The applied current on the SMA actuators (whether close
to or away from the rated current).
The cooling environment, (whether in a standstill air, fan
cooled environment or even immersed in a cooling fluid).
Fig. 10. The tactile display pin simulation behavior;
(a) Not actuated ES, SS (contracted & soft). (b) Not actuated ES
(contracted), actuated SS (hard). (c) Actuated ES (elongated), not
actuated SS (soft). (d) Actuated ES, SS (elongated and hard).
3-D Model Applied force
2 mm Diameter
Boundary
condition: Fixed
SS
ES
15
mm
25 m
m
187
V. CONCLUSIONS
The conceptual design of a multimodal tactile display device
for displaying both the shape and stiffness of an object was
presented. The tactile display device consists of two springs,
one for displaying the shape and the other for stiffness
display. The spring material is made from SMA to control its
shape and stiffness. In the system design, the selection of
spring specification to suit the stiffness range of human
tissue was considered.
In design, it was shown that by using a display pin of
2mm diameter, a tissue Young’s modulus ranged from
35 to 100 kPa can be displayed by spring stiffness of 85
to 250N/m, respectively.
The designed spring’s parameters are simulated for the
purpose of the design validation and the simulation
results showed that the proposed device can display
stiffness ranged from 95 to 275 N/m, independent of
shape displacement.
REFERENCES
[1] M. Nakatani, H. Kajimoto, D. Sekiguchi, N. Kawakami and S. Tachi,
“3D Form Display with Shape Memory Alloy”, ICAT Vol. 8, pp 179-
184, 2003.
[2] V.G. Chouvardas, A.N. Miliou, M.K. Hatalis, “Tactile displays:
Overview and recent advances”, Displays 29, 2008, pp. 185–194.
[3] N. Asamura, N. Yokoyama, H. Shinoda, A method of selective
stimulation to epidermal skin receptors for realistic touch feedback, in:
Proceedings IEEE Virtual Reality Conference, 1999, pp. 274–181.
[4] T. Maeno, K. Kobayashi, N. Yamazaki, Relationship between the
structure of human finger tissue and the location of tactile
receptors,Bulletin of JSME International 41 (1), 1998, pp. 94–100.
[5] M.V. Ottermo, Stavdahl, and R.A. Johansen. Electromechanical
design of a miniature tactile shape display for minimally invasive
surgery. In World Haptics, pp. 561-562, 2005
[6] A. Talbi, O. Ducloux, N. Tiercelin, Y. Deblock, P. Pernod, and V.
Preobrazhensky. Vibrotactile using micromachined electromagnetic
actuators array. In Actuator 2006, pp. 120-123, 2006.
[7] T. Debus, T.J. Jang, P. Dupont, and R.D. Howe. Multi-channel
vibrotactile display for teleoperated assembly. International Journal of
Control, Automation, and Systems, pp. 390-397, 2004.
[8] J. Pasquero and V. Hayward. Stress: a practical tactile display system
with one millimeter spatial resolution and 700 Hz refresh rate. In
Proceedings of Eurohaptics, pp. 94-110, 2003.
[9] Q. Wang and V. Hayward. Compact, portable, modular, high
performance, distributed tactile display device based on lateral skin
deformation. In 14th Symposium on Haptic Interfaces for Virtual
Environment and Teleoperator Systems IEEE, pp.67-72, 2006.
[10] P.M. Taylor, D.M. Pollet, A. Hosseini-Sianaki, and C.J. Varley.
Advances in an electrorheological fluid based tactile array. pp. 135-
141, 1998.
[11] P. Sommer-Larsen and R. Kornbluh. Overview and recent advances in
polymer actuators. In Actuator 2006, pp. 86-96, 2006.
[12] H. Herr and R. Kornbluh. New horizons for orthototic and prosthetic
technology: arti_cial muscle for ambulation. In Smart Structures and
Materials: Electroactive Polymer Actuators and Devices, 2004.
[13] Arabic site, http://www.tbeeb.net/a-1441.htm, (accessed in Jan 30
2012)
[14] W. C. Hayes, L. M. Keer, G. Herrmann, L. F. Mockros, "A
mathematical analysis for indentaion test of articular cartilage " J.
Biomech., vol. 5, pp. 541-551, 1972.
[15] K. Chinzei, K. Miller, K. Homma, K. Hyodo, "Experiments of Brain
Tissue Deformation; in vitro and in vivo, vol. 35, pp.316, 1997.
[16] M. Farshad, M. Barbezat, P. Flüeler, F. Schmidlin, P. Graber and P.
Niederer, "Material characterization of the pig kidney in relation with
the biomechanical analysis of renal trauma", Journal of Biomechanics,
vol. 32, no. 4, pp.417-425, 1999.
[17] J. Kim, B. K. Tay, N. Stylopoulos, D. W. Rattner and M. A.
Srinivasan, "Characterization of intraabdominal tissues from in vivo
animal experiments for surgical simulation," 6th International Medical
Image Computing and Computer Assisted Intervention (MICCAI)
Conference, Montreal, Canada, vol. 1, pp.206-213, 2003.
[18] Y. Zheng, and A.F.T Mak., 1999, "Effective Elastic Properties for
Lower Limb Soft Tissues from Manual Indentation Experiment,"
IEEE Transactions on Rehabilitation Engineering, Vol. 7, No. 3,
pp.257-267.
[19] K. J. Voss, M. A. Srinivasan, Investigation of the internal geometry
and mechanics of the human fingertip, in vivo, using magnetic
resonance imaging, Tech. Report, no. 622. Research Laboratory of
Electronics, Massachusetts Institute of Technology, 1998.
[20] M. Zhang, Y. P. Zheng and A. F. T. Mak, "Estimating the effective
Young's modulus of soft tissues from indentation tests-nonlinear finite
element analysis of effects of friction and large deformation," Med.
Eng. Phys., Vol. 19, no. 6, pp. 512-517, 1997.
[21] A. Falvo, “Thermomechanical characterization of Nickel-Titanium
Shape Memory Alloys,” PhD thesis, Dept. Mech. Eng., Della Calabria
Univ., 2008.
[22] R. O. Parmley, “Illustrated Sourcebook of Mechanical Components”,
Third Edit., McGraw-Hill, 2000.
188
