Ferroelectric Ceramics, Monte Verita, © Birkhliuser Verlag Basel 363

FERROELECTRIC SENSORS AND ACTUATORS:

SMART CERAMICS

Robert E. Newnham

"Smart" materials have the ability to perform both sensing and actuating functions.
Passively smart materials respond to external change in a useful manner without
assistance, while actively smart materials have a feedback loop which allows them to
both recognize the change and initiate an appropriate response through an actuator
circuit.
One of the techniques used to impart intelligence into materials is "Biomimetics," the
imitation of biological functions in engineering materials. Composite ferroelectrics
fashioned after the lateral line and swim bladders of fish are used to illustrate the idea.
"Very Smart" materials, in addition to sensing and actuating, have the ability to "learn"
by altering their property coefficents in response to the environment. Field-induced
changes in the nonlinear properties of relaxor ferroelectrics and soft rubber are utilized
to construct tunable transducers. Integration of multifunctional ferroic ceramics into
compact, robust packages is a major goal in the development of smart materials.

1 Introduction

It has been said that life itself is motion, from the single cell to the most complex organism: the
human body. This motion, in the form of mobility, change, and adaptation, is what elevates living
beings above the lifeless forms (Rogers, 1990). This concept of creating a higher form of
materials and structures by providing the necessary life functions of sensing, actuating, control,
and intelligence to those materials is the motivation for studying smart materials.
Smart materials are part of smart systems - functional materials for a variety of engineering
applications. Smart medical systems for the treatment of diabetes with blood sugar sensors and
insulin delivery pumps. Smart airplane wings that achieve greater fuel efficiency by altering their
shape in response to air pressure and flying speed. Smart toilets that analyze urine as an early
warning system for health problems. Smart structures in outer space incorporating vibration
cancellation systems that compensate for the absence of gravity and prevent metal fatigue. Smart
364 R. E. Newnham

toys like "Altered Beast" where one is awakened from the dead and must learn to survive in the
hostile environment of a different age. Smart houses with electrochromic windows that control the
flow of heat and light in response to weather changes and human activity. Smart tennis rackets
with rapid internal adjustments for overhead smashes and delicate drop shots. Smart muscle
implants made from rubbery gels that respond to electric fields, and smart dental braces made from
shape memory alloys. Smart hulls and propulsion systems for navy ships and submarines that
detect flow noise, remove turbulence, and prevent detection. Smart water purification systems that
sense and remove noxious pollutants. A number of smart systems have already been developed for
automobiles, but there are many more to come. In a recent newspaper cartoon, Blondie and
Dagwood encountered a smart automobile that drives itself back to the finance company when the
owner misses a payment!
In this chapter the idea of "smartness" in a material is discussed, along with a number of
examples involving ferroelectric components. Some of these smart ceramics are in production,
while others have great potential but are thus far limited to laboratory investigations.
To begin our discussion, we first define "smart" to set the limits for classifying smart
materials.

2 How Smart is Smart?

The short answer is "not very." Webster's dictionary gives several definitions for the word
SMART, including "alert, clever, capable," "stylish," and "to feel mental distress or irritation." All
three definitions are appropriate for the currently fashionable subject, "smart materials." They are
"stylish," they are - in some cases - "clever," and it does cause some of us "mental distress" to
think that a ceramic might somehow possess intelligence, even in rudimentary form.
There are many words in the English language denoting various degrees of intelligence.
Beginning at the bottom, an intelligence scale might look like this: stupid - dumb - foolish - trivial-
sensible - smart = clever - intelligent - wise. Many modem day materials have been cleverly
designed to carry out useful functions and, in some cases, that we are justified in calling them
"smart." They are decidedly better than "sensible" materials, but calling them "intelligent" seems
rather presumptuous and self-serving. Perhaps in the future, when we are able to integrate
information - processing and feedback circuitry into our sensor and actuator materials, perhaps
then we will be justified in calling our materials "intelligent." As pointed out later, such a time is
not far off.
To clarify the concept of smart materials, we describe a few examples of passive and active
smartness.
Ferroelectric Sensors and Actuators: Smart Ceramics 365

3 Passive Smartness

A passively smart material has the ability to respond to environmental conditions in a useful
manner. A passively smart material differs from an actively smart material in that there are no
external fields or forces or feedback systems used to enhance its behavior. The "S" words in
Table I summarize some of the meanings of passive smartness.

Table I: Some Attributes of Passive Smartness

Selectivity
Self Diagnosis
Self Tuning
Sensitivity
Shapeability
Self Recovery
Simplicity
Self Repair
Stability and Multistability
Stand-by Phenomena
Survivability
Switchability

Many passively smart materials incorporate self-repair mechanisms or stand-by phenomena which
enable the material to withstand sudden changes in the surroundings. The crack-arresting
mechanisms in partially stabilized zirconia are a good example. Here the tetragonal-monoclinic
phase change accompanied by ferroelastic twin wall motion are the stand-by phenomena capable of
generating compressive stresses at the crack tip. In a similar way, toughness can be improved by
fiber pull-out or by multiple crack-branching as in the structural composites used in aircraft, or in
machinable glass-ceramics.
Ceramic varistors and PTC thermistors are also passively smart materials. When struck by
high-voltage lightning, a zinc oxide varistor loses most of its electrical resistance and the current is
by-passed to ground. The resistance change is reversible and acts as a stand-by protection
phenomenon. Varistors also have a self-repair mechanism in which its highly nonlinear I-V
relationship can be restored by repeated application of voltage pulses. Barium titanate PTC
thermistors show a very large increase in electrical resistance at the ferroelectric phase
transformation near 130°C. The jump in resistance enables the thermistor to arrest current surges,
366 R. E. Newnham

again acting as a protection element. The R(V) behavior of the varistor and the R(T) behavior of
the PTe thermistor are both highly nonlinear effects which act as standby protection phenomena,
and make the ceramics smart in a passive mode.

4 Active Smartness

A smart ceramic can also be defined with reference to sensing and actuating functions, in analogy
to the human body. A smart ceramic senses a change in the environment, and using a feedback
system, makes a useful response. It is both a sensor and an actuator. Examples include vibration
damping systems for outer space platforms and electrically-controlled automobile suspension
systems using piezoelectric ceramic sensors and actuators.
The Piezoelectric Pachinko machine illustrates the principle of an actively smart material.
Pachinko Parlors with hundreds of vertical pinball machines are very popular in Japan. The
Piezoelectric Pachinko game constructed by engineers at Nippon Denso is made from PZT
multilayer stacks which act as both sensors and actuators. When a ball falls on the stack the force
of impact generates a piezoelectric voltage. Acting through a feedback system, the voltage pulse
triggers a response from the actuator stack. The stack expands rapidly throwing the ball out of the
hole, and the ball moves up a spiral ramp during a sequence of such events. Eventually, it falls
into a hole and begins the spiral climb allover again.
A video tape head positioner developed by Piezoelectric Products, Inc., operates on a similar
principle. A bilaminate bender made from tape-cast PZT ceramic has a segmented electrode pattern
dividing the sensing and actuating functions of the positioner. The voltage across the sensing
electrode is processed through the feedback system resulting in a voltage across the positioning
electrodes. This causes the cantilevered bimorph to bend, following the video tape track path.
Articulated sensing and positioning electrodes operating at 450Hz near the tape head help keep the
head perpendicular to the track.
These two examples illustrate how a smart ceramic operates. Both sensing and actuating
functions are involved in its performance, and while the Japanese scientists have a somewhat
different perspective on "smart" or "intelligent" materials (Takagi, 1989), the end results of efforts
in this area are very similar.

5 Rubber-Like Ceramics

Every baseball and cricket player knows the impOltance of "soft hands." In catching a baseball, it
is important to withdraw the hands slightly on making contact with the ball. This reduces the
Ferroelectric Sensors and Actuators: Smart Ceramics 367

momentum of the ball gradually and creates a soft landing. Soft landings are achieved on ceramics
in the same way, making them feel as soft as rubber.
To test the concept, controlled compliance experiments have been carried out using PZT
sensors and actuators (Newnham et aI., 1989). In the test set-up, one actuator is used as the
external driver, and the other as the responder. Sandwiched between the two actuator stacks are
two sensors and a layer of rubber. The upper actuator is driven at a frequency of 100 Hz and the
vibrations are monitored with the upper sensor. The pressure wave emanating from the driver
passes through the upper sensor and the rubber separator and impinges on the lower sensor. The
resulting signal is amplified using a low noise amplifier and fed back through a phase shifter to the
lower actuator to control the compliance.
A smart sensor-actuator system can mimic a very stiff solid or a very compliant rubber. This
can be done while retaining great strength under static loading, making the smart material
especially attractive for vibration control.
If the phase of the feedback voltage is adjusted to cause the responder to contract in length
rather than expand, the smart material mimics a very soft, compliant substance. This reduces the
force on the sensors and partially eliminates the reflected signal. The reduction in output signal of
the upper sensor is a measure of the effectiveness of the feedback system. In our experiments, the
compliance of the actuator-sensor composite was reduced by a factor of six compared to rubber
(Newnham et aI., 1989).

6 Modulated Suspension Systems

The automobile industry is a very large market in which smart composites and sensors are already
widely used. More than fifty electroceramic components can be found in today's high-tech autos,
ranging from the air-fuel oxygen sensors used in most autos to the more exotic piezoelectric
raindrop sensor, which automatically senses the amount of rain falling and adjusts the windshield
wipers to the optimum speed (Taguchi, 1987).
Controlled compliance with piezoelectric ceramics is utilized in Toyota's piezoTEMS (Toyota
Electronic Modulated Suspension), a system which has been developed to improve the drivability
and stability of the automobile, and at the same time enhance passenger comfort (Tsuka et al.,
1990). The TEMS is basically a road stability sensor and shock adjustor, which detects bumps,
dips, rough pavement, and sudden lurches by the vehicle, then rapidly adjusts the shock absorbers
to apply a softer or firmer damping force, depending on what is necessary to minimize discomfort
while maintaining control of the vehicle. The shock absorbers are continuously readjusted as the
road conditions change so that rocking or wobbling on soft shocks is eliminated.
368 R. E. Newnham

The TEMS road surface sensor consists of a five-layer piezoelectric ceramic sensor mounted
on the piston rod of the shock absorber. When a bump in the road is encountered, the resulting
stress applied to the sensor produces a voltage which is fed into an electronic control unit that
amplifies the signal and supplies a high voltage to the piezoelectric actuator. The 88-layer PZT
actuator produces a 50 11m displacement on the oil system which is hydraulically enlarged to two
millimeters, enough to change the damping force from firm to soft; the entire process takes only
about twenty milliseconds (not even enough time to slam on the brakes!) Also figured into the
actuator output are the vehicle speed and the driver's preference for a generally softer (American)
or firmer (European) ride.
Alternatively, it is possible to damp stresses and vibrations without the need for a sensor-
actuator feedback loop; materials which can perform this function are called passive damping
materials. In a piezoelectric passive damper, a piezoelectric ceramic is connected in parallel with a
properly matched resistor. The external stress creates a polarization in the piezoelectric, which
induces a current in the resistor, leading to energy dissipation. A high piezoelectric coupling
coefficient is required to induce the maximum voltage and energy dissipation Ramachandran et al.,
1990).

7 Actuator Materials

There are many approaches to controlling vibration and structural deformation. Actuation strain
can be controlled by piezoelectric materials (Burke and Hubbard, 1987), electrostrictive materials
(Uchino, 1986), magnetostrictive materials (Butler, 1988), shape-memory metal alloys
(Schetky,1979), and thermally-controllable materials (Edberg, 1987). Utilizing a system with
distributed actuators, it is possible to design structures with intrinsic vibration and shape control
capabilities. Among the most important actuator materials are shape memory metals and ceramics.
The shape memory effect is exhibited by alloys which undergo thermoelastic martensite
transformations. This is a first order displacive transformation in which a body centered cubic
metal transforms by shear on cooling to a martensitic phase. When deformed in the martensitic low
temperature phase, shape memory alloys will recover this deformation and return to the original
shape when heated to a temperature where the martensite reverts back to the parent body-centered
cubic structure. Unlike most ferroelectric and ferromagnetic transitions, the shape memory
transformation has a large hysteresis which can be troublesome in practice.
Alloys exhibiting the shape memory effect fall into two general classes: non-ferrous and
ferrous. Non-ferrous alloys currently in commercial use are Ni-Ti, Cu-Zn-Al, and Cu-Ni-Al.
Ferrous shape memory alloys under development include Fe-Pt, Fe-Ni-C and Fe-Ni-Co-Ti.
Ferroelectric Sensors and Actuators: Smart Ceramics 369

Non-ferrous shape memory alloys of nickel-titanium alloy (Nitinol) have been developed by
Goodyear Aerospace Corporation for spacecraft antennae (Schetky, 1979). A wire hemisphere of
the material is crumpled into a tight ball, less than five centimeters across. When heated above
77OC, the ball opens up into its original shape--a fully formed antenna. Although it has seldom
been used in service, this antenna demonstrates the magnitude of deformation and reformation
possible in shape memory alloys.
While shape memory alloys are more like a solution looking for a problem, it has been
suggested that transient and steady state vibration control can be accomplished with hybrid
structures in which the shape memory alloy is embedded inside the material (Rogers et al., 1989).
Some ceramic materials also possess a sizeable shape-memory effect; of particular interest are
materials which are simultaneously ferroelectric and ferroelastic. Their ferroelasticity ensures that
recoverable spontaneous strain is available for contributing to the shape memory effect, and the
ferroelectricity implies that their spontaneous strain can be manipulated not only by mechanical
forces but also by electric fields.
Shape-memory has been demonstrated in PLZT ceramics, an important ferroelectric-
ferroelastic because of the tremendous potential for applications due to the formation of
microdomains smaller than the wavelength of light. In one experiment, a 6.5/65/35 PLZT helix
was heated to 200°C, well above the transition temperature for recovery, TF (= TC>, then
mechanically loaded and cooled to 38°C (well below TF) - the "brittle" PLZT helix was deformed
by 30% after the load was removed. Upon heating to 180°C, above TF, the helix transformed
back to its original shape, dramatically demonstrating the shape memory effect in brittle ceramics
(Wadhawan et al., 1981).
Researchers at Sophia University in Tokyo have created a multilayer shape memory actuator
with a (Pb, Nb)(Zr, Sn, Ti)03 ceramic. Twenty rectangular plates of the ceramic were stacked in
an MLCC-type structure; the magnitude of the strains that were induced were small by comparison
to most shape memory alloys (3-4 microns) but were three times larger than the strains produced
using conventional piezoelectric actuators. In this case, a ferroelectric to antiferroelectric phase
change is responsible for the shape change (Furuta and Uchino, 1990).

8 Electrorheological Fluids

One of the criteria which separates "smart" materials from "very smart" or "intelligent" materials is
the ability of the material to not only sense a change and actuate a response, but to automatically
modify one or more of its property coefficients during the sensing/actuating process. In effect, this
type of material not only warns the user of a change in its environmental conditions and responds
to it, but can in addition adjust itself to compensate for future change.
370 R. E. Newnham

Electrorheological (ER) fluids (Ghandi and Thompson, 1989) and their magnetic analog,
ferrofluids (Rosenweg, 1985), are an example of materials that have great potential for use in
smart materials and systems. ER fluids are typically suspensions of fine particles in a liquid
medium; the viscosity of the suspension can be changed dramatically by applying an electrical
field. The electric field causes alignment of the particles in fibril-like branches in the direction of
the applied field. The alignment disappears when the electric field is removed, thus creating the
desired property of complete cyclic reproducibility.
ER fluids represent an advanced class of composite materials with self-tuning properties, that
will find considerable use in vibration control applications. In addition, the compatibility of this
technology with modem solid state electronics makes it an attractive component for integration into
multifunction, self-contained smart material packages.

9 Biomimetics· Fish Ears

The word "biomimetic" is not found in most dictionaries so it needs to be defined. It comes from
the Greek words "bios," meaning "life," and "mimetikos," meaning "to imitate." Biomimetic
means to imitate life, or to use the biological world as a source of ideas for device concepts.
Fish and the other inhabitants of the underwater world have some interesting ways of talking
and listening which have been copied using piezoelectric ceramics. For most
fish, the principal sensors are the lateral line and the inner ear coupled to the swim bladder. The
pulsating swim bladder also acts as a voice, as do chattering teeth in certain fish species.
The lateral line runs fom the head to the tail of the fish and resembles a towed array with
sensing organs (stitches) spaced at intervals along the nerve fiber. Each stitch contains several
neuromasts made up of gelatinous cupulae resembling pimples in shape, within each cupula are
several fibers which vibrate as the fish swims through water and acts as sensors for flow noise.
The hair-like fibers are extremely thin in diameter ranging from 0.5 to 10 ~. When stimulated by
turbulence, the motion of the hairs produces changes in the synapses which are in tum connected
to the nerve fiber. The electric signal originates from impedance changes in cell walls which
modulate the flow of K+ ions. The lateral line is especially sensitive to low frequency fluid motion
parallel to the length of the fish. In the 50 Hz range, threshold signals are observed for
displacements as small as 30 nm ! (Bond, 1979; Moyle and Cech, 1988).
The 1-3 composite hydrophones described later are patterned after the hair-ftlled cupulae of
the lateral line. Thin PZT fibers embedded in polymer provide excellent electromechanical coupling
to a liquid medium and can be used as both sensors and actuators.
Ferroelectric Sensors and Actuators: Smart Ceramics 371

10 Dr. Dolittle and Fish Talk

Among the most popular children's books of all time are the Doctor Dolittle books written by
Hugh Lofting. Doctor Dolittle could talk to the animals. He started with "Pig" and went on to
"Duck" and "Cow", and eventually mastered 498 languages. There is a marvelous scene in the
movie where young Tommy Stebbins meets Dr. Dolittle in his laboratory and watches him talk to
goldfish by blowing bubbles in the fish bowl through a rubber tube. Between bursts of bubbles,
Dr. Doolittle listened intently to through a stethoscope pressed against the side of the bowl.
In recent years, great advances have been made in recording and understanding fish talk,
largely because of the development of improved hydrophone arrays and high speed spectrum
analyses. Much of the talk is in the form of low frequency grunts below 200Hz. Sound functions
in a variety of ways for fish, both in offense and in defense, for warning and intimidation. Many
fish speak differently during breeding season, and appear to use coded repetition rates to
communicate. Our ability to "farm the oceans" could be greatly enhanced by learning how to talk
to fish and control their movements and feeding habits.
Although they do not possess a larynx, many species of fish produce high-pitched sound by
grinding their teeth, but the vibration of the swim bladder wall provides the greatest repertoire of
noises or calls. The croakers of Chesapeake Bay make tapping noises like a woodpecker, by
contracting their drumming muscles attached to the swim bladder, and the twilight choruses of sea
robins caused great confusion among the operators of antisubmarine echo-location devices during
World War II (Urick, 1975).

11 Inner Ears and Swim Bladders

The nature of sound transmission in water has had a great influence on the evolution of hearing in
fish. Sound, especially low frequency sound, travels faster and farther than in air. "Near-field"
sound consists of small fluid motions or vibrations and are characterized by a displacement
direction. They are detected by the inner ear or by the lateral line. The hydrostatic component or
"far-field" sound is detected best through the swim bladder.
The inner ear is made up of inertia-sensing chambers resembling accelerometers. Within each
chamber is a dense ear stone (otolith) which vibrates in a near-field sound wave. The inertia of the
ear stone causes it to lag behind the motion of the fish, and to push against hair cells which line the
chamber (sacculus). On bending, the hair cellular membranes deform, stimulating neural
transmissions to the brain. Connections to the swim bladder improve the sensitivity to far-field
sound.
372 R. E. Newnham

The primary function of the gas-filled swim bladder is to provide buoyancy, but it is also used
for sound and pressure reception and in some species is equipped with drumming muscles for
sound production. The flexible swim bladder responds to hydrostatic pressure waves by changing
volume. Fish with swim bladders can perceive relative pressure changes equivalent to less than
0.5% of the ambient hydrostatic pressure.
Direct or indirect linkages from the swim bladder to the inner ear promote the hearing sensation.
Fish with no connections perceive low frequency sound (less than 500 Hz), while those with good
connections have an upper frequency response of 5000 Hz. As might be expected, the swim
bladder is reduced in size with depth, and loses much of its sensitivity as a sensor.

12 Hydrophone Materials

The knowledge which comes from the understanding of "fish talk" can be directly applied to
research in materials destined to someday "sleep with the fishes". Hydrophones are underwater
listening devices made from piezoelectric materials which respond to hydrostatic pressure waves.
Among the applications for hydrophones are sonar systems for submarines, off-shore oil
platforms, geophysical prospecting equipment, fish finders, and earthquake monitors.
As the earth's population continues to increase, mankind must continue to search for new and
efficient sources of food and nutrition. The world's oceans may provide a solution to this
problem, not only through fish farming but through the use of new and varied salt water
vegetation that could provide an abundant source of food, especially for third world countries in
which poor soil and harsh climates prohibit conventional farming. Smart hydrophone transceivers
will receive and transmit fish talk and monitor the growth of underwater vegetation.
The figure of merit for hydrophone materials is the product of the hydrostatic piezoelectric
charge coefficient (db) and the piezoelectric voltage coefficient (gb). While good piezoelectric
materials such as PZT have high d33 and d31 piezoelectric coupling coefficients, the db value is
only about 45 pC/N because d33 and d31 are opposite in sign, and db = d33 + 2d31. dbgb is also
inversely related to the dielectric permittivity, E33, so that low dielectric constants are desirable as
well.
Rather than abandon PZT in search of the ultimate hydrophone material, avoiding this
problem is often a matter of appropriate engineering of existing materials. Too often in the field of
materials research we put too much emphasis on the synthesis of new materials and too little
emphasis on new and unique designs for old materials.
A composite design with 1-3 connectivity is similar in design to the hair-filled gelatinous
cupula with thin PZT rods embedded in a polymer matrix. The 1-3 piezocomposites have excellent
sensitivity to pressure waves in water (Klicker et al., 1981; Gururaja et al., 1984). The large d33 is
Ferroelectric Sensors and Actuators: Smart Ceramics 373

maintained because the parallel connection results in stress transfer from polymer to piezoceramic,
while the d31 is destroyed because of series connection in the lateral dimension where the
mechanical load is absorbed by the polymer and not transferred to the PZT rods. Finally, £33 is
minimized due to the large volume of low £33 polymer present. The dhIDl values are improved by
more than an order of magnitude when very thin rods are used.
Another piezoelectric hydrophone composite maximizes dh by simply redirecting the applied
stresses using specially shaped electrodes (Xu et al., in press). These are flextensional transducers
which mimic the motions of the swim bladder. Shallow air spaces are positioned under the metal
electrodes while the PZT ceramic plays the role of the muscle lining the swim bladder. When
subjected to a hydrostatic stress, the thick metallic electrodes convert a portion of the z-direction
stress into large radial and tangential stresses of opposite signs. The result is that d31 changes from
negative to positive, so that its contribution now adds to d33 rather than subtracting from it. The
dhgh of these composites is approximately 250 times that of pure PZT.

13 Very Smart Composites: the Tunable Transducer

By building in a learning function, the defmition of a smart material can be extended to a higher
level of intelligence: A very smart material senses a change in the environment and responds by
changing one or more of its property coefficients. Such a material can tune its sensor and actuator
functions in time and space to optimize future behavior. With the help a feedback system, a very
smart material becomes smarter with age, something even human beings strive for. The distinction
between smart and very smart materials is essentially one between linear and nonlinear properties.
The physical properties of nonlinear materials can be adjusted by bias fields or forces to control
response.
To illustrate the concept of a very smart material, we describe the tunable transducer recently
developed in our laboratory. Electromechanical transducers are used as fish fmders, gas igniters,
ink jets, micropositioners, biomedical scanners, piezoelectric transformers and filters,
accelerometers, and motors.
Five important properties of a transducer are the resonant frequency fr, the acoustic
impedance ZA, the mechanical damping coefficient Q, the electromechanical coupling factor k, and
the electrical impedance ZE. The resonant frequency and acoustic impedance are controlled by the
elastic constants and density, as discussed in the next section. The mechanical Q is governed by
the damping coefficient (a.) and is important because it controls "ringing" in the transducer.
Definitions of the coefficients are given in Table II. Electromechanical coupling coefficients are
controlled by the piezoelectric coefficient which, in tum, can be controlled and fine-tuned using
relax or ferroelectrics with large electrostrictive effects. The dielectric "constant" of relaxor
374 R. E. Newnham

felToelectrics depends markedly on DC bias fields, allowing the electrical impedance to be tuned
over a wide range as well. In the following sections we describe the nature of nonlinearity and
how it controls the properties of a tunable transducer.

Table II: Important characteristics of an electromechanical transducer.

Fundamental resonant where

frequency,f,ofthe
thickness mode
f=;t~ t = thickness dimension,
c = elastic stiffness,
p = density
where
Acoustic Impedance ZA IZAI=~ c = elastic stiffness,
p = density
where
Mechanical Q Q=-
7t A.a = acoustic wavelength,
A. ·a a = damping coefficient
Electromechanical coupling where
coefticient k
k=d~ d = piezoelectric charge coefficient,
e = electric permittivity

Electrical impedance Ze where

~
IZ.I = ill. . A ill =
angular frequency,
A = electrode area

14 Elastic Nonlinearity: Tuning the Resonant Frequency

Infonnation is transmitted on electromagnetic waves in two ways: amplitude modulation (AM) and
frequency modulation (PM). There are a number of advantages to FM signal processing,
especially where lower noise levels are important. Atmospheric static is considerably lower in PM
radio than in AM.
Signal-to-noise ratios are also important in the ultrasonic systems used in biomedical and
nondestructive testing systems, but PM-modulation is difficult because resonant frequencies are
controlled by stiffness (c) and transducer dimensions (t). Neither c, t, nor the density (p) can be
tuned significantly in ceramics and most other materials, but rubber is an exception. To tune the
resonant frequency of a piezoelectric transducer, we designed and built a composite transducer
incorporating thin rubber layers exhibiting nonlinear elasticity (Xu et al., 1990).
Ferroelectric Sensors and Actuators: Smart Ceramics 375

Rubber is a highly nonlinear elastic medium. In the unstressed compliant state, the molecules
are coiled and tangled, but under stress the molecules align and the material stiffens noticeably.
Experiments carried out on rubber-metal laminates demonstrate the size of the nonlinearity.
Young's modulus (E = 1/s1111) was measured for a multilayer laminate consisting of alternating
steel shim and soft rubber layers each 0.1 mm thick. Under compressive stresses of 200 MPa, the
stiffness is quadrupled from about 600 to 2400 MPa (Rivin, 1983). The resonant frequency f is
therefore doubled, and can be modulated by applied stress.
Rubber, like most elastomers, is not piezoelectric. To take advantage of the elastic
nonlinearity, it is therefore necessary to construct a composite transducer consisting of a
piezoelectric ceramic (PZT) transducer, thin rubber layers, and metal head and tail masses, all held
together by a stress bolt.
The resonant frequency and mechanical Q of such a sandwich structure has been measured as
a function of stress bias. Stresses ranged from 20 to 100 MPa in the experiments. Under these
conditions the radial resonant frequency changed from 19 to 37 kHz, approximately doubling in
frequency as predicted from the elastic nonlinearity. At the same time the mechanical Q increases
from about 11 to 34 as the rubber stiffens under stress.
The changes in resonant frequency and Q can be modeled with an equivalent circuit in which
the compliance of the thin, rubber layers are represented as capacitors coupling together the larger
masses (represented as inductors) of the PZT transducer and the metal head and tail masses. Under
low stress bias, the rubber is very compliant and effectively isolates the PZT transducer from the
head and tail masses. At very high stress, the rubber stiffens and tightly couples the metal end
pieces to the resonating PZT ceramic. For intermediate stresses the rubber acts as an impedance
transformer giving parallel resonance of the PZT - rubber - metal -radiation load.
Continuing the biomimetic theme, it is interesting to compare the change in frequency of the
tunable transducer with the transceiver systems used in the biological world. The biosonar system
of the flying bat is similar in frequency and tunability to our tunable transducer. The bat emits
chirps at 30 kHz and listens for the return signal to locate flying insects. To help it differentiate the
retUl11 signal from the outgoing chirp, and to help in timing the echo, the bat puts an PM signature
on the pulse. This causes the resonant frequency to decrease from 30 to 20 kHz near the end of
each chirp. Return signals from the insect target are detected in the ears of the bat where neural
cavities tuned to this frequency range measure the time delay and flutter needed to locate and
identify its prey. Extension of the bat biosonar principle to automotive, industrial, medical and
enteltainment systems is underway.
376 R. E. Newnham

15 Piezoelectric Nonlinearity: Tuning the Electromechanical Cou-

pling Coefficient

The difference between a smart and a very smart material can be illustrated with piezoelectric and
electrostrictive ceramics. PZT (lead zirconate titanate) is a piezoelectric ceramic in which the
ferroelectric domains have been aligned in a very large poling field. Strain is linearly proportional
to electric field in a fully poled piezoelectric material which means that the piezoelectric coefficient
is a constant and cannot be electrically tuned with a bias field. Nevertheless it is a smart material
because it can be used both as a sensor and an actuator.
PMN (lead magnesium niobate) is not piezoelectric at room temperature because its Curie
temperature lies near O°C. Because of the proximity of the ferroelectric phase transformation,
however, and because of its diffuse nature, PMN ceramics exhibit very large electrostrictive
effects.
Electromechanical strains comparable to PZf can be obtained with electrostrictive ceramics
like PMN, and without the troubling hysteretic behavior shown by PZf under high fields. The
nonlinear relation between strain and electric field in electrostrictive transducers can be used to tune
the piezoelectric coefficient and the dielectric constant
The piezoelectric d33 coefficient is the slope of the strain-electric field curve when strain is
measured in the same direction as the applied field. Its value for Pb(MgO.3Nbo.6Tio.l) 03 ceramics
is zero at zero field and increases to a maximum value of 1300 pCIN (about three times larger than
PZT) under a bias field of 3.7 kV/cm .
This means that the electromechanical coupling coefficient can be tuned over a very wide
range, changing the transducer from inactive to extremely active. The dielectric constant also
depends on DC bias. The polarization saturates under high fields causing decreases of 100% or
more in the capacitance. In this way the electrical impedance can be controlled as well.
Electrostrictive transducers have already been used in a number of applications including
adaptive optic systems, scanning tunneling microscopes, and precision micropositioners (Uchino,
1986).
To summarize, two types of nonlinearity are utilized in the fully tunable transducer: elastic
nonlinearity and piezoelectric nonlinearity. By incorporating thin rubber layers in an electrostrictive
transducer several important properties can be optimized with bias fields and bias stresses.
Electromechanical coupling coefficients and electric impedance are tuned with electric field, and
mechanical damping, resonant frequency, and acoustic impedance with stress bias.
Ferroelectric Sensors and Actuators: Smart Ceramics 377

16 Smart Electroceramic Packages

Up to this point, our discussion has focused primarily on piezoelectric transducers in which the
sensing and actuating functions are electromechanical in nature. But the idea of a smart material is
much more general than that. There are many types of sensors and many types of actuators, and
many different feedback circuits.
Many of these sensors and actuators can be fabricated in the form of multilayer ceramic
packages. Until recently multilayer packages consisted of low permittivity dielectric layers with
metal circuitry printed on each layer and interconnected through metallized via holes between
layers. Buried capacitors and resistors have now been added to the three-dimensional packages,
and other components will follow shortly. Smart sensors, adaptive actuators, and display panels,
together with thermistors and varistors to guard against current and voltage overloads, are next in
line for development (Newnham, 1988).
Integration and miniaturization of electroceramic sensors and actuators is an ongoing process
in the automotive and consumer electronics areas. Multilayer packages containing signal
processing layers made up of low-permittivity dielectrics and printed metal interconnections are in
widespread production. Further integration with embedded resistors and capacitors are under
development, and its seems likely that intelligent systems will make use of the same processing
technology. Tape casting and screen printing are used most often. Varistors, chemical sensors,
thermistors, and piezoelectric transducers can all be fabricated in this way, opening up the
possibility of multicomponent, multifunction ceramics with both sensor and actuator capabilities.
Silicon chips can be mounted on these multifunctional packages to provide all or part of the control
network. Processing is a major challenge because of the high firing temperatures of most
ceramics, typically in the range 8000 C to 1500oC. Differences in densification shrinkage and
thelmal contraction, together with adverse chemical reactions between the electroceramic phases,
create formidable problems. Nevertheless, the rewards for such an achievement are substantial. An
all-ceramic multifunction package would be small, robust, inexpensive, and sufficiently refractory
to withstand elevated temperatures.
Electrodes are both a problem and a challenge. At present, precious metals such as palladium
and platinum are used in multilayer ceramic components, greatly adding to the cost of the device.
Copper and nickel electrodes require that the ceramic be fired in a reducing atmosphere, which
may reduce the electroceramic layers and adversely affect the electrical properties. Copper and
silver have high electrical conductivity but the melting points (-1000°C) require lower firing
temperatures and make it necessary to alter the ceramic compositions and fabrication procedures.
Some headway has been made on this problem, but further work is needed. One interesting
approach to the problem is ceramic electrodes. There are a number of ceramic phases with
378 R. E. Newnham

excellent conductivity which could be used, including the copper-oxide superconductors. In

actuator devices, there are some special advantages in having electrodes and piezoceramics with
matched elastic properties.
Composites are another approach to making sensor-actuator combinations. These can be
formed at lower temperatures using low-firing ceramics and high temperature polymers such as
polyimides. Sol-gel and chemical precipitation methods are helpful in preparing ceramic powders
with low calcining temperatures, but further work on composite fabrication is required to obtain
reliable and reproducible electrical behavior.
To miniaturize the sensors and actuators, and to obtain complex shapes, we recommend the
use of photolithography and other processing methods employed in the semiconductor industry.
Ultraviolet curable polymers incorporated into the tape-casting process make photolithographic
processing comparatively easy and should find wide use in preparing ceramic or composite
packages for intelligent systems.
The next logical step is to combine the sensor and actuator functions with the control system.
This can be done by depositing electroceramic coatings on integrated circuit silicon chips.
Reliability is a major requirement in all complex systems. Further research on electrical and
mechanical breakdown of sensor-actuator materials is needed to elucidate the mechanisms
responsible for failure. Fail-safe and self-repair phenomena would be very helpful.
Intelligent systems for hostile environments is another topic area of interest. Sensor and
actuator systems that can operate at high temperatures inside engines or furnaces are required to
monitor combustion and to provide pollution controls. Radioactive and chemical waste sensors are
another problem. Intelligent systems for oceanographic studies and oil exploration must withstand
high pressures and salinity conditions.
Electroceramics have a vital role to play in intelligent systems, and many new developments
will take place in the coming decade and the next century.

References

Bond, C.E., 1979: "Biology of Fishes," Saunders College Publishing Co., Philadelphia.

Burke, S., and J.E.Hubbard, 1987: "Active Vibration Control of a Simply-Supported Beam
Using a Spatially-Distributed Actuator," IEEE Control Systems Magazine 7,25-30.

Butler,I.L., 1988: "Application Manual for Design of Magnetostrictive Transducers," Edge

Technologies, Extrema Division, N.Marshfield, MA.
Ferroelectric Sensors and Actuators: Smart Ceramics 379

Edberg,D.L., 1987: "Control of Flexible Structures by Applied Thermal Gradients," AIAA

Journal, 25, 877-883.

Furuta,A., K.Oh, and K.Uchino, 1990: "Mechanical Clamper Using Shape Memory Ceramics,"
presented at the 7th International Symposium on Applications for Ferroelectrics, June 6-8,
Urbana, Illinois.

Ghandi, M.V., and B.S.Thompson, 1989: in "Smart Materials, Structures, and Mathematical
Issues," ed. by C.A.Rogers, Technomic Publishers, Lancaster, PA, 63-68.

Gururaja, T.R., W.A.Schulze, L.E.Cross, R.E.Newnham, B.A.Auld, and J.Wang, 1984:

"Resonant Modes in Piezoelectric PZT Rod-Polymer Composite Materials," Proc. of IEEE
Ultrasonic Symposium, 523-527.

Klicker, K.A., J.V.Biggers, and R.E.Newnham, 1981: Composites of PZT and Epoxy for
Hydrostatic Transducer Applications," J.Amer.Cer.Soc., 64, 5-9.

Moyle, P.B., and J.J.Cech, Jr., 1988: "Fishes: An Introduction to Ichthyology," Prentice Hall,
Englewood Cliffs, N.J..

Newnham, R. E., 1988: "The Golden Age of Electroceramics," Adv. Ceram. Mat., 3,12-16.

Newnham, R.E., Q.C.Xu, S.Kumar, and L.E.Cross, 1989: "Smart Ceramics," J.Wave-Matl.
Int., 4, 3-10.

Ramachandran,A.R., Q.C.Xu, L.E.Cross, and R.E.Newnham, 1990: "Passive Damping with

Piezoelectric Ceramics," presented at the 7th International Symposium on Applications for
Ferroelectrics, June 6-8, Urbana, illinois.

Rivin, E.I., 1983: "Properties and Prospective Applications of Ultra Thin Layered Rubber-Metal
Laminates for Limited Travel Bearings," Tribology Int., 16, 17-25.

Rogers, C.A., 1990: "From the Editor" Journal of Intelligent Material Systems and Structures, I,
3.
Rogers, C.A., C.Liang, and D.K.Barker, 1989: in "Smart Materials, Structures, and
Mathemetical Issues," ed.by C.A.Rogers, Technomic Publishers, Lancaster, PA, 39-62.

Rosenweig, R.E., 1985: "Ferrohydrodynamics," Cambridge University Press, New York.

Schetky,L., 1979: "Shape Memory Alloys," Scientific American, 241, 74-82.

Taguchi,M., 1987: "Applications of High-Technology Ceramics in Japanese Automobiles," Adv.

Ceram. Mat., 2, 754-762.

Takagi, T., 1989: "The Concept oflntelligent Materials and the Guidelines on R&D Promotion",
Japan Science and Technology Agency Report.
380 R.E.Newnham

Tsuka,H., J.Nakomo, and Y.Yokoya, 1990: "A New Electronic Controlled Suspension Using
Piezoelectric Ceramics," IEEE Workshop on Electronic Applications in Transportation.

Uchino,K., 1986: "PiezoelectriclElectrostrictive Actuators," Morikita Publishers, Tokyo.

Urick, R.J., 1975: "Principles of Underwater Sound," McGraw Hill, New York.

Wadhawan, V.K., M.C.Kernion, T.Kimura, and R.E.Newnham, 1981: "The Shape Memory
Effect in PLZT Ceramics," Ferroelectrics, 37, 575-578.

XU, Q.C., J.Belsick, S.Yoshikawa, T.T.Srinivasan, and R.E.Newnham, in press: "Piezoelectric

Composites with High Sensitivity and High Capacitance for Use at High Pressures," IEEE
Proceedings.

XU, Q.C., R.E.Newnham, M.Blaskiewicz, T.T.Fang, T.T.Srinivasan, and S.Yoshikawa, 1990:

"Nonlinear Multilayer Composite Transducers," Presented at the 7th International Symposium
on Applications for Ferroelectrics, June 6-8, Urbana, lllinois.

Robert E. Newnham, Materials Research Laboratory, The Pennsylvania State University,

University Park, PA 16802, USA