See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245430868

Smart Spring Control of Vibration on Helicopter Rotor Blades

Article in Journal of Aircraft · March 2009

DOI: 10.2514/1.27914

CITATIONS READS

24 718

3 authors:

Gregory Oxley F. Nitzsche

Envision Energy Carleton University
9 PUBLICATIONS 148 CITATIONS 147 PUBLICATIONS 1,808 CITATIONS

SEE PROFILE SEE PROFILE

Daniel Feszty
Carleton University
36 PUBLICATIONS 860 CITATIONS

SEE PROFILE

All content following this page was uploaded by F. Nitzsche on 10 September 2014.

The user has requested enhancement of the downloaded file.

JOURNAL OF AIRCRAFT
Vol. 46, No. 2, March–April 2009

Engineering Notes
ENGINEERING NOTES are short manuscripts describing new developments or important results of a preliminary nature. These Notes should not exceed 2500
words (where a figure or table counts as 200 words). Following informal review by the Editors, they may be published within a few months of the date of receipt.
Style requirements are the same as for regular contributions (see inside back cover).

Smart Spring Control of Vibration Active control of aeroelastic response and helicopter blade
vibration has been a matter of great interest in the helicopter industry
on Helicopter Rotor Blades and the scientific community in general. In particular, individual
blade control (IBC) [1] has gained special attention with the
development of actuators capable of performing satisfactorily in a
Gregory Oxley,∗ Fred Nitzsche,† and Daniel Feszty‡ rotating frame reaching a significant degree of maturity. In recent
Carleton University, Ottawa, Ontario K1S 5B6, Canada years, several actuation schemes have been proposed, most of them
DOI: 10.2514/1.27914 based on solid-state actuators [2]. It is still a challenge, however, to
produce solid-state actuators that can provide satisfactory results for
full-scale helicopters. The main limitation is generally recognized to
Nomenclature be the lack of sufficient control actuation power when full-scale
aerodynamic and inertial loads are present. The piezoelectric
c1 , c2 = Smart Spring viscous damping coefficients
materials, from which most solid-state actuators are built, are
associated with primary and secondary load paths
characterized by large specific energy. However, this energy is not
F = external (input) force applied to the Smart Spring
readily available at the frequencies requested by IBC. Recent results
k = effective dynamic stiffness of the Smart Spring
indicate that active composite fibers (ACF) made from piezoelectric
k1 , k2 = Smart Spring constants associated with primary and
material embedded in host composite matrices may produce enough
secondary load paths
strain near the structural resonance, reaching the threshold of
m = effective inertia of the Smart Spring
applicability in full-scaled blades [3]. The IBC concept using
m1 , m2 = mass of external and internal sleeves in the Smart
piezoelectric stacked material in combination with a mechanic
Spring
amplification device was used to activate a flap successfully flown in
N = contact force applied by piezoelectric stack
a full-scale rotor [4].
T = Smart Spring period of actuation
Smart materials are characterized by a limited stroke, although
t = time
x = displacement (output) yielded by the Smart Spring they are able to generate a considerably large force. To circumvent
y = displacement associated with the Smart Spring this problem, the use of “smart” springs to indirectly achieve control
secondary load path over the vibration problem encountered in helicopter rotors has been

= dynamic stiffness complex coefficients proposed [5]. The Smart Spring is best suited for use in actively
 = dry friction coefficient altering the boundary conditions of a beam structure and, as such, it

= Smart Spring control frequency, 2=T should be incorporated at the root of the blade acting as a mechanical
! = Smart Spring frequency of excitation filter to attenuate the vibrations induced by the unsteady
aerodynamic loads passing to the rotor hub. Although it can be
used in any rotor configuration (hinged or hingeless) alone, it is better
Superscripts
to use it in an IBC system cooperating with another flow-control
_ = time derivative device such as a servo flap or an ACF-embedded blade [6]. European

= complex conjugate and U.S. patents were granted to the National Research Council of
Canada (NRCC) for the Smart Spring [7], and experiments in both
still air and a wind tunnel for a nonrotating, full-scale blade were
I. Introduction conducted at NRCC with a prototype actuator to verify its
performance [8]. An independent investigation demonstrated that
H ELICOPTERS generate significant vibration and noise, which
deteriorates their operational economics and performance. The
main sources of vibration are the rotor hub reactions of the inertial
small changes in the nominal stiffness of the blade root are sufficient
to attenuate all vibration loads up to a 90% efficiency [9]. This is not
and aerodynamic loads on the rotor blades. In a rotor composed of an unexpected result because modifying the boundary conditions
identical blades, the rotor hub cancels most of the vibratory loads has, in general, a significant effect on the vibration characteristics of a
except for the harmonic multiples of the number of blades. mechanical system.
The present paper explores the Smart Spring as a means of
Presented as Paper 2270 at the 46th AIAA/ASME/ASCE/AHS/ASC attenuating helicopter airframe vibration. This is achieved by
Structures, Structural Dynamics and Materials Conference, Austin, Texas, modeling the Smart Spring as a replacement for the conventional
18–21 April 2005; received 21 September 2006; revision received 18 October rotor pitch link connecting the swash plate and blade pitch horn in
2008; accepted for publication 19 October 2008. Copyright © 2008 by the SMARTROTOR, an advanced aeroelastic/aeroacoustic simulation
American Institute of Aeronautics and Astronautics, Inc. All rights reserved. tool for actively controlled rotors developed in a partnership
Copies of this paper may be made for personal or internal use, on condition involving Carleton University, the National Technical University of
that the copier pay the $10.00 per-copy fee to the Copyright Clearance Center, Athens in Greece, and the Massachusetts Institute of Technology in
Inc., 222 Rosewood Drive, Danvers, MA 01923; include the code 0021-8669/ the United States [10]. In the present application, the structural model
09 $10.00 in correspondence with the CCC.
∗ is based on the classical linear three-dimensional Euler beam theory
Research Assistant, Department of Mechanical and Aerospace
Engineering. Student Member AIAA.
attributable to Houbolt and Brooks [11], enhanced to include rigid-
†
Professor, Department of Mechanical and Aerospace Engineering; body modes due to the blade articulation. The aerodynamic model
[email protected]. Senior Member AIAA. employs a discrete vortex-particle description of the flow that takes
‡ into consideration the free-wake characteristics, coupled with a panel
Assistant Professor, Department of Mechanical and Aerospace
Engineering. Member AIAA. method representation of the rotor blades [12]. In SMARTROTOR,
692
 J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES 693

the Farassat 1A formulation of the Ffowcs–Williams and It is worthwhile to stress that the variation on the dynamic stiffness
Hawkings’s equation generates the aeroacoustic results after the parameter is directly proportional to the electrical stimulus, which is
pressure distribution on the blades is obtained [13]. SMARTROTOR externally applied to the piezoelectric stack. In the simpler case, the
was validated against the experimental results produced by the control law is harmonic and, as such, it can be expanded into a
European HELINOISE program in a previous publication [14]. complex Fourier series:

II. The Smart Spring Principle X

r1 Z T
The Smart Spring principle is summarized in Fig. 1, in which two kt 
r eir
t
r  1=T kteir
t dt
springs, k1 and k2 , are associated with two load paths (primary and r1 0 (2)
secondary, respectively). External and internal sleeves of masses m1 r  1; 2; . . . ; 1
and m2 , respectively, compose the assembly on the right-hand side of
the figure. The latter sleeves can slide with respect to each other,
preventing the two springs from being connected. An external force,
As discussed earlier, the coefficients in Eq. (2) are best obtained
Ft, is applied to the external sleeve, which is attached to the spring
from experimental characterization. The analytical solution of this
designated by k1 . A stack of piezoelectric actuators is inserted into
dynamic problem is well known because it represents a linear
the internal sleeve, which is attached to the spring designated by k2 .
differential equation with periodic coefficients, which is solved by
When the actuators are off, the two sleeves can move freely and the
the Floquet method [15]. Another solution to the same problem was
resulting displacement is xmax  F=k1 . When the actuators are
presented in a previous work using the harmonic balance method and
turned on, the stack compresses the internal sleeve by a normal force,
Hill’s infinite determinants [16]. The latter solution is particularly
N. Only when the resultant friction force, N, applied by the internal
interesting to pursue here because it helps to understand the proposed
sleeve on the external sleeve is sufficiently large and the two sleeves
form of control. It is assumed that the solution is harmonic on the
are forced together is an arrangement of two springs in parallel
exciting frequency:
obtained. The resulting displacement decreases to its minimum,
xmin  F=k1 
k2 . Therefore, the stiffness of the system rises from its original X
1
value, k1 , to the final value, k1  k2 . In fact, the spring designated by xt  xr eir!t (3)
k2 is forced into motion by the resultant friction force applied by the r1
internal on the external sleeve, which can be controlled by the
external electrical stimulus. The displacement of the system under
the external force varies between the two extremes, xmin and xmax . In The coefficients associated with the negative harmonics are
general, depending on the dimensions and manufacturing tolerances simply the complex conjugates of their positive harmonic
of the system, maximum displacement may not even be achievable counterparts (xr  x
r ). For the sake of this demonstration, the
because the maximum stroke supplied by the stack of piezoelectric control law will be composed by only the first harmonic. If the time-
elements might be insufficient to guarantee that the two sleeves move averaged value of the dynamic stiffness coefficient is
0 , one has
freely in the actuators-off condition. kt 

1 ei
t 
0 
1 ei
t because
1 

1 . Substituting the
When motion is involved during the actuation cycle defined by the latter expression for kt into Eq. (1), using Eq. (3), and setting the
electric stimulus, the friction coefficient varies between its highest control frequency equal to the exciting frequency, ! 
,
(static) and its lowest dynamic limits. The actual variation process
may be complicated, and the problem is more suitable to be treated
experimentally rather than analytically. As an approximation, one X
1

can assume a process in which dry friction or “structural” damping is mr2 !2 
0 eir!t 

1 eir1!t 
1 eir1!t xr  Fei!t
present. In fact, one may not expect that a perfect contact between the r1

two sleeves will ever occur; therefore, some slippage always exists. (4)
This slippage, however, is beneficial to the robustness of the system
because otherwise an on–off type of control would be present,
implying that during the transients many blade structural modes (This is actually the case for most applications involving helicopter
could be excited. IBC.)
The Smart Spring is a single-input, single-output system (SISO) in A transformation in the dummy indices of the summation is
which the total displacement x (output) “seen” by the external performed next. In Eq. (4), the two terms involving frequency shifts
harmonic force F (input) at any time is governed by Eq. (1), are rewritten as
mx  ktx  Fei!t (1)
X
1 X
1 X
1
assuming that the inertia of the SISO system is represented by xt  xr eir 1!t
 xr 1 eir!t  xr 1 eir!t (5)
m  m1  m2 and the magnitude of the resulting dynamic stiffness r1 r 11 r1
parameter varies between the two static limits, k1  jktj 
k1  k2 , according to a control law that may or may not be harmonic.
Replacing the latter series into Eq. (4), the following is obtained:

X
1
mr2 !2 
0 xr 

1 xr1 
1 xr1 eir!t  Fei!t  Fei!t
r1
(6)

In Eq. (6), the harmonic exciting force was expanded to include its
mirror image, negative harmonic, to keep the series real valued
(F1  F1  F). Applying harmonic balance, the following set of
tridiagonal equations on the coefficients of the frequency response is
Fig. 1 Smart Spring principle. obtained:
694 J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES

2 3
..

the transmitted higher-harmonic loads (i.e., the ones generated by the
6 .
1 7 aerodynamic excitation of the torsion structural mode of the blade in
6 7 this case, which is known to lie in the range of 3–5=rev for a typical
6
1 4m!2 
0

7
6 1 7 helicopter blade and, therefore, would be critical in a four-blade
6 7
6
1 2

m! 
0
1 7 rotor) from the blade to the swash plate. As also suggested by the
6 7
6
7 diagram of the Smart Spring in Fig. 1, the equations that govern the
6
1
0
1 7
6 7 motion are
6 2
7
6
1 m! 
0
1 7
6 7
6 2
7 m1 x  c1 x_  k1 x  Ft tNt
6
1 4m! 
0
1 7
(9)
4 5
.. m2 y  c2 y_  k2 y  tNt

1 .
8 9 8 9
> .
.. > > .. > where viscous damping coefficients are now introduced in the
>
> >
> >
> .>>
>
> >
> >
> >
> primary and secondary load paths, respectively. The upper or lower
>
> >
> >
> >
> x2 >
> > > >0> >
> signs of the dry friction terms are used if x_ > y_ or x_ < y,
_ respectively,
>
> >
> >
> >
>
>
> >
> >
> > and t is a function of the absolute relative velocity between the
> x1 >
> > > >F> >
< = < > = sleeves, jx_  yj,
_ and the contact materials properties. This model has
x0  0 (7) been extensively studied, and its validity was checked against the
>
> >
> >
> >
> hardware experimental results in another publication [18]. When the
>
> >
> >
> >
>
>
> x1 >> >
>F> > piezoelectric stack actuation force is off, Nt  0, Eq. (9) is
>
> >
> > > > >
>
> >
> >
> >
>
>
>
x2 >>
>
>
>
> 0> >
>
decoupled and the external load is transmitted solely through the
>
> > > > > > primary load path. When the actuation force is on, the external load is
: .. > ; : .. > ;
. . shared between the primary and secondary load paths.
In the present work, the Smart Spring has been implemented as a
An approximate solution for the lower harmonics of xt is subloop of the main aeroelastic calculations in SMARTROTOR
achieved by successively increasing the number of the harmonics (Fig. 2). It provides the boundary condition for the pitch degree of
retained. The numerical convergence is very fast, even if F is freedom of the structural module. In the main aeroelastic loop of
composed of a large number of harmonics. This is due to the narrow SMARTROTOR, the aerodynamic module calculates the blade
band nature of the system of equations. The band of the system is loading, determines the rigid-body motion of the blade, and feeds
twice plus one the number of harmonics present in the control law (a these results back to the aerodynamic module. This is done iteratively
control law with two harmonics would produce a system of band until convergence is reached. When the Smart Spring is introduced as
five). However, if the control law itself has fast convergence a subloop in the main aeroelastic loop, the pitch link loads are passed
properties, that is, the relative importance of its higher harmonics is as an input to the Smart Spring in the form of the external load. The
increasingly less (a fair assumption in any real case), the numerical discrete version of Eq. (9) is solved at each time step for the output
convergence of the solution will remain fast. The first approximation displacement using an implicit Newmark algorithm. The change in
for the solution in the present case involves solving a 3 3 set of the pitch due to the output displacement of the Smart Spring is then
equations on the harmonic coefficients with indices of 1, 0, and 1, fed back to the structural module.
respectively: The control strategy implemented in this study is on–off and based
on the goal of maximum energy extraction from the system [19]. At

1 

1  each time step, the relative velocities of m1 and m2 are examined and
x1  x
1 x0  F it is determined whether a control action will promote the motion of
m! 
0 
0  2
1

1
2
(8) m1 , thereby adding energy to the system, or whether it will hinder the
m!2 
0 
0 
21 
1

1 motion of m1 , thereby extracting energy from the system. This is
x1  F
m!2 
0 m!2 
0 
0  2
1

1 done by verifying the sign of the product y _ x_  y.
_ If it is positive (or
zero), the control system maintains Nt  0. Otherwise, if negative,
A set of discrete complex frequency response functions between
the input excitation F and the output displacement x is obtained. The
introduction of the Smart Spring caused an active redistribution of
the dynamic response spectrum, whose first approximation is given
by Eq. (8). Therefore, helicopter IBC using the Smart Spring concept
is achieved by the spectral redistribution of the unsteady
aerodynamic loads. Because in the present demonstration no
viscous damping was included, dissipation is solely due to hysteretic
damping.
As one can infer from this discussion, the present concept
fundamentally differs from other concepts devised to control
aeroelastic response and to perform IBC [17]. In the presented
concept, the energy supplied by the piezoelectric elements is used to
do indirect rather than direct work against the external forces. This is
a general principle that is expected to reduce the demand of power
that is normally imposed on piezoelectric and other smart materials at
frequencies at which they are not well suited for use. It is worthwhile
to point out that this type of use of the smart material is more
demanding on the force rather than stroke supplied by the material,
which certainly favors the piezoelectric inherent characteristics.

III. Smart Spring Model and Control Algorithm

Figure 1 represents the physical model of a linear Smart Spring
proposed to replace the helicopter pitch link (connecting the root
blade to the swash plate) to actively reshape the spectrum of some of Fig. 2 SMARTROTOR code flowchart with Smart Spring device.
 J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES 695

Table 1 Simulation Parameters

Blade geometry Simulation parameters Smart Spring constants
No. blades 4 Time step 72=rev m1 0.500 kg
Blade planform Rectangular Rotor speed 109:22 rad=s m2 0.050 kg
Rotor radius 1.85 m Advance ratio 0.25 c1 1000 Ns=m
Chord length 0.121 m Thrust trim 3875 N c2 200 Ns=m
Twist distribution Linear (4 to 2 deg) Moment trim 0 Nm k1 5:0 106 N=m
Airfoil type NACA 23012 Revolutions 6 k2 241 N=m
Nmax 200 N

the control system sets Nt  Nmax , the maximum possible value for Spring acts as a filter for the higher harmonics of the axial load
the actuation force. vibration spectrum, and it does not affect the harmonics associated
with the rotor trim and cyclic control (0 and 1=rev, respectively)
when its internal resonance is set at 4=rev. The small differences seen
IV. Results and Analysis at 1 and 2=rev are due to numerical errors. The FFT analysis also
The test case in which the Smart Spring will be demonstrated as a indicates that harmonics higher than 2=rev were either attenuated or
vibration attenuation device is based on the BO105 rotor. Table 1 not considerably affected by the Smart Spring actuation. Significant
summarizes the simulation parameters used in this analysis. The reductions in the critical 4 and 8=rev (those multiples of the number
Smart Spring parameters were tuned to achieve the control strategy of rotor blades) were verified.
(maximum power extraction) at the frequency of 4=rev, which is
considered critical for a four-blade rotor. Hence, in the primary load
path, a linear spring of very high constant k1 was set to guarantee the V. Conclusions
original stiffness of the pitch link and, thereof, the blade stability. The Numerical simulation results indicated that the Smart Spring is a
secondary load path was tuned to resonate at the 4=rev frequency. structural control device able to attenuate helicopter rotor higher-
Figure 3 depicts the axial load transmitted through the pitch link harmonic loads transmitted to the fuselage through the pitch link. The
during one revolution of the rotor as predicted by SMARTROTOR Smart Spring does not affect the rotor trim, collective and cyclic
(after six complete revolutions to achieve the steady-state condition). control commands if it is tuned to internally resonate at the rotor
The value at 0=rev is the magnitude of the time-averaged force higher harmonic frequencies that are the objective of control. The
experienced by the pitch link. This value is the same for the force transmitted through the pitch link with the Smart Spring was
controlled and uncontrolled situations, indicating that the Smart verified to be 44% of the original value at 4=rev and 62% of the
Spring, in fact, redistributes the vibration spectrum (arbitrarily original value at 8=rev, which may be considered modest compared
truncated at 10=rev in the figure). The figure suggests that the Smart with other active vibration reduction approaches used in rotorcraft.

Fig. 3 Time history of the axial load transmitted through the pitch link with (dashed line) and without (solid line) the Smart Spring active control and
corresponding FFT of the same data.
696 J. AIRCRAFT, VOL. 46, NO. 2: ENGINEERING NOTES

Acknowledgments 166, 2000.

[10] Cesnik, C. E. S., Opoku, D. G., Nitzsche, F., and Cheng, T., “Active
The authors would like to thank S. Voutsinas and V. Riziotis from Twist Rotor Blade Modeling Using Particle-Wake Aerodynamics and
the National Technical University of Athens for useful discussions Geometrically Exact Beam Structural Dynamics,” Journal of Fluids
and technical support for this project. and Structures, Vol. 19, pp. 651–668.
doi:10.1016/j.jfluidstructs.2004.01.007
[11] Houbolt, J. C., and Brooks, G. W., “Differential Equations of Motion
References for Combined Flapwise Bending,” Chordwise Bending, and Torsion of
[1] Kretz, M., and Larche, M., “Future of Helicopter Rotor Control,” Twisted Nonuniform Rotor Blades, NACA Rept. 1346, 1958.
Vertica, Vol. 4, pp. 13–22, 1980. [12] Voutsinas, S. G., and Triantos, D. G., “High-Resolution Aerodynamic
[2] Friedmann, P. P., and Millott, T. A., “Vibration Reduction in Rotorcraft Analysis of Full Helicopter Configurations,” American Helicopter
Using Active Control: A Comparison of Various Approaches,” Journal Society Paper C11, Sept. 1999.
of Guidance, Control, and Dynamics, Vol. 18, No. 4, pp. 664–673. [13] Ffowcs-Williams, J. E., and Hawkings, D. L., “Sound Generation by
doi:10.2514/3.21445 Turbulence and Surfaces in Arbitrary Motion,” Philosophical
[3] Wilbur, M. L., Yeager, W. T., and Sekula, M. K., “Further Examination Transactions of the Royal Society of London, Series A: Mathematical
of the Vibratory Loads Reduction Results From The NASA/Army/MIT and Physical Sciences, Vol. A264, pp. 321–342, 1969.
Active Twist Rotor Test,” American Helicopter Society 58th Annual [14] Nitzsche, F., and Opoku, D. G., “Acoustic Validation of a New Code
Forum, Royal Aeronautical Society, Bristol, England, U.K., 2002, Using Particle Wake Aerodynamics and Geometrically Exact Beam
pp. 60.1–60.15. Structural Dynamics,” The Aeronautical Journal, Vol. 109, No. 1096,
[4] Jänker, P., Hermle, F., Lorkowski, T., Storm, S., and Wettemann, M., June 2005, pp. 257–267.
“Advanced Piezoelectric Servo Flap System for Rotor Active Control,” [15] Dugundji, J., and Wendell, J. H., “Some Analysis Methods for Rotating
National Aerospace Laboratory Paper AR-10, Sept. 2006. Systems with Periodic Coefficients,” AIAA Journal, Vol. 21, pp. 890–
[5] Nitzsche, F., Lammering, R., and Breitbach, E., “Can Smart Materials 897, 1983.
Modify the Blade Root Boundary Conditions to Attenuate Helicopter doi:10.2514/3.8167
Vibration?,” Fourth International Conference on Adaptive Structures, [16] Nitzsche, F., “Aeroelastic Analysis of a Helicopter Blade with Active
Technomic Publishing, 1994, pp. 139–150. Impedance Control at the Root,” Canadian Aeronautics and Space
[6] Nitzsche, F., Feszty, D., Waechter, D., Bianchi, E., Voutsinas, S., Journal, Vol. 47, March 2001, pp. 7–16.
Gennaraetti, M., Coppotelli, G., and Ghiringelli, G. L., “The SHARCS [17] Jacklin, S. A., Haber, A., deSimone, G., Norman, T. R., Kitaplioglu, C.,
Project: Smart Hybrid Active Rotor Control System for Noise and and Shinoda, P., “Full-Scale Wind Tunnel Test of an Individual Blade
Vibration Attenuation of Helicopter Rotor Blades,” American Control System for a UH-60 Helicopter,” American Helicopter Society
Helicopter Society Paper 52, Sept. 2005. 58th Annual Forum, American Helicopter Society, Alexandria, VA,
[7] Nitzsche, F., Grewal, A., and Zimcik, D., “Structural Component June 2002.
Having Means for Actively Varying its Stiffness to Control [18] Harold, T., and Nitzsche, F., “The Smart Spring—An Indirect Active
Vibrations,” U.S. Patent No. 5,973,440, October 1999. Low-Power Adaptive Structure Device for Vibration Attenuation,”
[8] Yong, C., Zimick, D. G., Wickramasinghe, V., and Nitzsche, F., High Performance Structures and Materials II, edited by C. A. Brebbia
“Development of the Smart Spring for Active Vibration Control of and W. P. DeWilde, WIT Press, Southampton, England, U.K., 2004,
Helicopter Blades,” Journal of Intelligent Material Systems and pp. 297–306.
Structures,” Vol. 15, No. 1, pp. 37–47. [19] Nitzsche, F., Harold, T., Wickramasinghe, V. K., Young, C., and
doi:10.1177/1045389X04039655 Zimcik, D. G., “Development of a Maximum Energy Extraction
[9] Anusonti-Inthra, P., and Ghandi, F., “Helicopter Vibration Reduction Control for the Smart Spring,” Journal of Intelligent Material Systems
Through Cyclic Variations in Blade Root Stiffness,” Journal of and Structures, Vol. 16, No. 11/12, pp. 1057–1066.
Intelligent Material Systems and Structures, Vol. 11, No. 2, pp. 153–

View publication stats