Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Copy 14 of 35 copies
AD-A255 959liii111111 1 111 IlI~lil11 l iii
IDA DOCUMENT D- 1189
SDIO WORKSHOP ON PIEZOELECTRIC CERAMICACTUATORS FOR SPACE APPLICATIONS
Janet Sater
DTICM!1ECTESEP 2 41992
June 1992
92-25745
Prepaird forStrategic Defense Initiative Organization
Approved for public release; distribution unlimited
* AINSTITUTE FOR DEFENSE ANALYSES
180 1 N. Beauregard Street, Alexandria. Virginia 22311-1772
IDA Log No. HO 92-41982
DEFINITIONSIDA publishes the following documents to report the results of its work.
RepoReports are the most authoritative and most carefully considered products IDA publishes.They normally embody results of major projects which (a) have a direct bearing on
decisions affecting major programs, (b) address issues of significant concern to theExecutive Branch, the Congress and/or the public, or (c) address issues that havesignificant economic implications. IDA Reports are reviewed by outside panels of expertsto ensure their high quality and relevance to the problems studied, and they are releasedby the President of IDA.
Group Reports
Group Reports record the findings and results of IDA established working groups andpanels composed of senior individuals addressing major issues which otherwise would bethe subject of an IDA Report. IDA Group Reports are reviewed by the senior individualsresponsible for the project and others as selected by IDA to ensure their high quality andrelevance to the problems studied, and are released by the President of IDA.
PaersPapers, also authoritative and carefully considered products of IDA, address studies thatare narrower in scope than those covered in Reports. IDA Papers are reviewed to ensurethat they meet the high standards expected of refereed papers in professional journals orformal Agency reports.
DocumentsIDA Documents are used for the convenience of the sponsors or the analysts (a) to recordsubstantive work done in quick reaction studies, (b) to record the proceedings ofconferences and meetings, (c) to make available preliminary and tentative results ofanalyses, (d) to record data developed In the course of an investigation, or (e) to forwardinformation that is essentially unanalyzed and unevaluated. The review of IDA Documents
is suited to their content and intended use.
I The work reported in this document was conducted under contract MDA 903 89 C 0003 forthe Department of Defense. The publication of this IDA document does not indicateendorsement by the Department of Defense, nor should the contents be construed asreflecting the official position of that Agency.
REPORT DOCUMENTATION PAGE Forrm ApprovedIOMB No. 0704-0188
Pubc Reporong turden o Owe edlehon of ,or.emaon tS estmeted In avag I hour ps respone. ociuding the ore 6 non g insuoorS. eanhig exesng datra s. gaherng and ma ag the date needed. and€ompoelng end .r.emg 1te colocon of .mfonabon Send oitments regardog ths burde eSotet or any otoh aspect of hs odlacton ot ofntomason. mdudng auggecos s to, or ang thot burden. to WalhintonH=6gq147,ms Servics. Drectorate tor Intonmat Opolrbone and R.por. 1215 Jefferson Dews Highrsy. Suite 1204. Arngton. VA 222024302. and to t Ofoo* of Management and Budget. Papoww k RoJCwOn Pnyect
1. AGENCY USE ONLY (Leave blank) 12. REPORT DATE 3. REPORT TYPE AND DATES COVERED
I June 1992 Final--January - May 19924. TITLE AND SUBTITLE 5. FUNDING NUMBERS
SDIO Workshop on Piezoelectric Ceramic Actuators for Space C - MDA 903 89 C 0003Applications
T - T-R2-597.096. AUTHOR(S)
Janet Sater
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
REPORT NUMBER
Institute for Defense Analyses1801 N. Beauregard St. IDA Document D-1 189Alexandria, VA 22311-1772
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORINGAGENCY REPORT NUMBERSDIO/TN I
The Pentagon, Room 1 E167Washington, DC 20301
11. SUPPLEMENTARY NOTES
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
Approved for public release: distribution unlimited.
13. ABSTRACT (Maximum 200 words)
Lt. Col. Michael Obal of the SDIO Materials & Structures Office sponsored this workshop on piezoceramicactuators for space applications. Its purpose was threefold: to promote communications among theappropriate groups; to focus and prioritize issues for near-term improvements in actuator performance; and toidentify and suggest research plans to obtain such improvements. Presentations were given byrepresentatives from each of three invited groups: end users/designers; materials researchers; andmaterials/device manufacturers. Issues identified in the workshop seemed to fall into three major categories:(1) piezoceramic material performance limitations; (2) distribution of properties of vendor-supplied materials;and (3) actuator applications. Performance limitation issues include, for example, d31 vs. d33 performance,
displacement/volt, fatigue life at high strains, and tensile strain-to-failure. Property distribution issues includecomposition, capacitance, and surface finish. Actuator application issues include ceramic material/actuatorfabrication, material handling, application of electrodes to actuator, and post-composite fabrication actuatorperformance.
14. SUBJECTTERMS 15. NUMBER OF PAGESpiezoelectric materials, actuators, piezoelectric ceramic actuators, piezoceramic 283actuators, PZT, PLZT, PMN, vibration suppression, vibration 16. PRICE CODE
17. SECURITY CLASSIFIC, 'nCM 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACTOF REPORT OF THIS PAGE OF ABSTRACT
UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED SARNSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)
Pm-a*odby Af Sid ZWIS2WI.02
IDA DOCUMENT D- 1189
SDIO WORKSHOP ON PIEZOELECTRIC CERAMICACTUATORS FOR SPACE APPLICATIONS
Janet Sater
June 1992
Approved for public release; distribution unlimited
IDAINSTITUTE FOR DEFENSE ANALYSES
Contract MDA 903 89 C 0003Task T-R2-597.09
PREFACE
Lt.Col. Michael Obal of the Strategic Defense Initiative Organization, Materials andStructures Office, manages a wide variety of advanced technology programs addressing
needs for various systems. Programs in the area of adaptive structures, particularly for
space systems, have recently been initiated. Vibration suppression, one of the important
program areas, is a key feature in improving the target tracking and hit-to-kill performanceof the various systems. Results from some of the materials and structures vibration
suppression projects led to a desire to bring together the device end users, the materialsresearchers, and the materials/device manufacturers for discussion. The agenda was put
together by Lt.Col. Michael Obal, Dr. Fred Kahn (NRL), and Dr. Janet M. Sater (IDA) to
address the identified issues.
The workshop was hosted by IDA on February 25, 1992. IDA was requested
under SDIO Task T-R2-597.09 to participate in the workshop and to prepare a proceedings
to document the content of the workshop. This effort was subsequently carried out byDr. Janet Sater with input from Lt.Col. Michael Obal and Dr. Fred Kahn.
Accession For
NTIS GRA&IDTIC TAB 0Unannounced -
DTIC QUALITY INSPECTED 3 Just fication
By-Distribution/Availability Codes
Avail and/oriiiDiet Special
.1 1
CONTENTS
Preface .......................................................................... iii
Glossary ....................................................................... Vu
1. INTRODUCTION........................................................ 1-1
2. SPACECRAFr DESIGNERS AND USERS OFPIEZOCERAMIC ACTUATORS .......................................... 2-1A. Dr. Allen Bronowicki (TRW)...........................................2-1B.- Dr. Dean Jacot (Boeing) ............................................... 2-4C. Dr. Ben Wada (JPL)................................................... 2-5D. Dr. John Breakwell (Lockheed)......................................... 2-6E. Mr. Richard Alexius (Martin Marietta-Orlando).......................... 2-7F.- Dr. Robert Glaser (JPL)............................................... 2-8G. Dr. Vijay Varadan (Penn State) ......................................... 2-9
3. PIEZOELECMRC MATERIALS RESEARCHIERS ANDACTUATOR EXPERTS.................................................. 3-1A. Dr. Eric Cross (Penn State) ............................................ 3-1B. Dr. Stephen Winzer (Martin Marietta) .................................. 3-3C. Dr. Gene Haertling (Clemson)..........................................3-4D. Dr. Robert Newnham (Penn State)...................................... 3-5E. Mr. Frank A. Tito (NUWC)............................................ 3-6
4. MATERIALS AND DEVICE MANUFACTURERS.........................4-1A. Mr. Barry Koepke (Ailiant Techsystems) ............................... 4-1B.- Mr. Craig Near (Vernitron Piezoelectric Division) ....................... 4-2C. Mr. Melvin Main (Edo Corporation)..................................... 4-4D. Dr. John Galvani (AVX Ceramics)...................................... 4-4E. Mr. Leslie Bowen (Materials Systems) ................................. 4-5
5. DISCUSSION............................................................5-1A.- Material Improvements ................................................ 5-1B. Manufacturing Process Improvements.................................. 5-2C. Testing, Nondestructive Evaluation (NDE).............................. 5-3
v
Appendix A-- Adaptive Structures Programs for the Strategic Defense Initiative
Organization
Appendix B-- Workshop Agenda and List of Attendees 0
Appendix C-- Advanced Piezoelectric Ceramic Actuator Materials for Space Applications
Appendix D-- ACESA, ACTEX and AMASS PZT Material Needs
Appendix E-- High Strain Actuators 0
Appendix F-- Improved PZT Through Controlled Powder Synthesis and Bulk PowderProcessing
0
0
0
0
vi
0
GLOSSARY
ACESA Advanced Composites with Embedded Sensors and Actuators
ACTEX Advanced Control Technology Experiment
Ag silver
AMASS Advanced Materials Application to Space Structures
ASTREX Advanced Space Structures Technology Research Experiments
BT barium-titanate
FACT Fast Acting Control Thruster
Gr/Ep graphite/epoxy
GrITP graphite/thermoplastic
Hz hertz
IRAD internal research and development
J joules
JPL Jet Propulsion Laboratory
K degrees Kelvin
kHz kilohertz
m meters
M&S Materials and Structures
PM microns
rail thousandths of an inch
rmn millimeters
MV million volts
N newtons
vii
NDE nondestructive evaluation
NUWC Naval Undersea Warfare Center 0
ONR Office of Naval Research
Pt platinum
PLZT lead-lanthanum-zirconate-titanate
PMN lead-magnesium-niobate
psi pounds per square inch
PT lead titanate
PVDF poly vinylidene fluoride
PZT lead-zirconate-titanate
RT room temperature •
SDIO Strategic Defense Initiative Organization
SMA shape memory alloys
SPICE Space Integrated Controls Experiment
V volts
W watts
viii
1. INTRODUCTION
Lt.Col. Michael Obal of the Strategic Defense Initiative Organization, Materials andStructures Office, manages a wide variety of advanced technology programs. Programs in
the area of adaptive structures, particularly for space systems, have recently been initiated(Appendix A, pp. A-i to A-10). These programs are beginning to address solutions forpotential difficulties resulting from the following: absence of scheduled maintenance;limited or no ability to assess performance capability after time in orbit; and limited abilityto adjust performance capability. Potential benefits of the adaptive structures technologyinclude on-orbit system health monitoring and reporting, threat attack warning andassessment, and improved target tracking and hit-to-kill performance.
Vibration suppression is a key feature in improving the target tracking and hit-to-kill
performance of the SDIO space systems (Appendix A, pp. A-3 to A-4). One of the earlymaterials and structures (M&S) efforts in the area of vibration suppression was theAdvanced Composites with Embedded Sensors and Actuators program (ACESA). In thecourse of this program a number of issues associated with the use of piezoelectric ceramicdevices for both sensors and actuators were identified. Results from ACESA coupled withthose from other programs led to a workshop that brought together the device end users,the materials researchers, and the materials/device manufacturers.
This workshop was held at the Institute for Defense Analyses on February 25,1992. An agenda and list of attendees can be found in Appendix B. Copies of charts usedby each briefer can be found in Appendixes C through F. The format of this document
follows the organization of the agenda.
Lt.Col. Michael Obal, Program Manager, opened the meeting by describing theoverall M&S Adaptive Structures program. This also provided an introduction to the
1-I
workshop. The focus of the workshop, piezoelectric ceramic materials for space-based
actuators, was emphasized (Appendix C). Issues associated with these materials and
devices that have been identified by the spacecraft/end user community were briefly
discussed. These include, among others, the following:
1. Inherent piezoceramic material performance limitations such as fatigue life athigh strains;
2. Noted spreads/distributions in properties of vendor-supplied materials such ascapacitance and piezoelectric activity composition as well as geometrictolerances; and
3. Device fabrication-related difficulties such as material handling, electrical
connections, and electrical insulation. 0
Objectives of the workshop we-e clearly defined:
I. To enhance communication between space-based actuator users and thematerials research and the materials/device manufacturing communities;
02. To focus and prioritize issues for near-term improvements in actuator
performance; and
3. To identify and suggest research plans to obtain such improvements.
Dr. Fred Kahn reiterated these objectives, stating that information exchange 0(particularly of unmet needs) among the participants was critical. More productive
collaboration could be expected as a result
The remainder of the report is divided into four chapters. The first three chapters
correspond to the three groups represented at the workshop. Te fourth chapter includes S
the final discussion and summary.
1 Hereafter, "piezoelectric ceramic" is shortened to "piezoceramic."
1-2
2. SPACECRAFT DESIGNERS AND USERS OF
PIEZOCERAMIC ACTUATORS
The spacecraft designer community was asked to cover five major topical areas, as
follows:
I. The nature and .,sign parameters of present actuator applications;
2. Identification of actuator issues experienced in past applications;
3. Spacecraft qualification requirements that the actuator would have to meet inorder to be used in real systems, including ambient and life requirements;
4. Current and predicted actuator performance and reliability requirements, basedon structural design or on the number of cycles required at design or maximumactuator strain levels; and
5. Quantification of the "value added" to various space surveillance andinterceptor systems via implementation of advanced actuator materials andconfigurations.
Briefing charts used bI each of the speakers in this sessioa can be found in Appendix D.
A. DR. ALLEN BRONOWICKI (TRW)
TRVV ias been a significant participant in several SDIO programs involvingthe embedding of piezoelectric actuators in composites: Advanced Composites withEmbedded Sensors and Actuators (ACESA), Advanced Control Technology Experiment(ACTEX), and Advanced Materials Application to Space Structures (AMASS). TheACESA struts, 16 feet long, contaiin 24 sensors and 12 actuators each. The ACTEXtripod struts are shorter and coutain 8 sensors and 4 actuators each. For ACTEX, anadaptive structures flight experiment, it is believed that the space environment may affectlong-term performance of these devices. The environments of concern include thetemperature range from -250 *F to +250 *F and the associated thermal cycling, as well as
radiation
1 Refer to Appendix A. pages A-3 to A-4, for a description of these programs.
2-1
Dr. Bronowicki summarized the results of validation tests of Navy Type I and Type
II lead-zirconate-titanate (PZT) actuators for ACESA and AMASS (Appendix D, p. D-3).
Type 1 (PZT-4) actuators have a low intrinsic hysteresis and good stiffness match to the 0
graphite fibers in a composite but Type II (PZT-5A) actuators have a higher free strain.
Hysteresis in the Type 1 was reduced dramatically after it had been embedded in graphite/
epoxy (Gr/Ep). The tensile strain-to-failure 2 of the Type I in Gr/Ep at room temperature
(RT) was -600 micro-strain (g-strain). For the Type II in the same material it was 2,000 to 0
2,500 g-strain. However, in graphite/thermoplastic (Gr/TP) composites, the Type II could
only withstand 500 to 1,000 tensile p-strain. Compressive strain-to-failure was above
8,000 t-strain. Fatigue did not appear to affect stiffness. As long as p-strains during
fatigue cycles were kept below 600 for the Type I in Gr/Ep and 1,500 for the Type II in 0
Gr/Ep no actuation losses were observed. Thermal cycling, however, resulted in
10 to 30 percent reductions in actuation strain. This is believed to be due to high stresses
resulting from thermal expansion mismatch. When asked if the actuator was debonding or
actually breaking, Dr. Bronowicki replied that cracks were observed in the piezoceramic. 0
He indicated that the dynamic transfer functions were found to be quite insensitive to
temperature, though the Type II were more sensitive to temperature than the Type I.
Additional results can be found in the paper "Mechanical Validation of Smart Structures,"
pages D-7 to D-10. 0
Figures of merit were also described (p. D-4): strain coefficient (d31) or
charge/stress coefficient (d3lo 1), lateral stress coefficient 3 (yE 1 1d31 ) or charge/strain
coefficient4 (YEId31j 1 ), and the planar coupling coefficient 5 (iKp). The lateral stress
coefficient is constant for most PZT piezoceramics at -10 N/V-n. It is desirable thatyE1 ld3 1 be flat with temperature. TRW would like manufacturers to measure the lateral
stress coefficient and/or the charge/strain coefficient. Vernitron measures lcp, which differs
from the charge/strain coefficient by the square root of the dielectric constant. 0
TRW has come up with a preliminary specification for PZT actuators and/or
sensors. Dr. Bronowicki solicited verbal or written suggestions from workshop attendees
02 Failure is defined as "irrecoverable degradation in actuation constant Ed31." Ed31 is a material property
that is used to optimize both sensing and actuation.3 This is an indicator of ability to actuate against a stiff structure. A larger number is better.4 This is an indicator of ability to sense lateral strain using a charge amplifier.5 This is similar to the lateral stress coefficient and is a measure of conversion efficiency between 0
electrical and mechanical energy.
2-2
0
for improvements in the specification (p. D-5, pp. D-I 1 to D-15). Basic requirements for
Type I and Type IT include the following:
1. Piezoceramic wafer thicknesses 50.02 inches with length and width controlledwithin +/-0.02 in.;
2. Electrode thickness = 0.0002 in. of nickel;
3. Wafers capable of withstanding 2 psi loads on a flat surface;6
4. Measured capacitance for each wafer within 5 percent of lot mean;7
5. Lateral stress output coefficient at RT (YIld31 ) 2 10 N/V-m;8 and
6. Planar coupling coefficient 9 (nominal at RT, icp) > 0.55 with a variation innominal Kip within +/- 10 percent of RT value over a temperature range of-125 °C to +125 OC.
Dr. Bronowicki also presented a "wish list" for PZT materials and devices (p. D-6):
thinner, multilayer wafers (-2.5 mils thickness per layer); cylindrically curved wafers at
-2.5 mils thickness per layer,10 wafers that operate in the d33 mode rather than the d31
mode; wafers that operate in the d15 mode in the plane of the composite for torsion; and
materials that have 10 times the stroke of PZTs with no loss of linearity or bandwidth and
have 1,000 to 2,000 p-strain actuation capability. One of the attendees indicated that
Motorola has a proprietary method for slicing PZTs down to a thickness of 2.5 mils for a
speaker application. These sliced wafers can be bent a significant amount, dependent
somewhat on grain size. Actuation could be increased as much as 2.5 times by using thed33 mode rather than the d31 mode. This requires a thin stack 1 of PZTs in the plane of the
composite, a very difficult manufacturing task. Hagood has devaloped a technique thatuses alternating electrode fingers to approximate this type of behavior: the PZT thicicness is
5 mils and the alternating electrodes are placed 30 mils apart.
6 Surface flatness appears to be a continuing problem.
7 Typically, there is a 10 percent variance in capacitance.8 Measured values are typically 20 to 30 percent lower than the book value.
9 TRW would like to have this number measured for each lot. Low temperature sensitivity of this valueis also desirable to ensure the gain margin.
10 The diameter of curvature is I in.
11 The thickness perpendicular to the stack is 5 mils. Each layer of the stack is 2.5 mils thick. SeeAppendix D, page D-6.
2-3
B. DR. DEAN JACOT (BOEING)
Boeing has been involved in a number of programs in the "Smart Structures" area
for space. Included among these are embedding of piezoceramics in Gr/Ep composites,
damping of aluminum via piezoceramics, ACESA (Phase 1), and embedding of
piezoceramics for damping trusses, among others (p. D-17). Dr. Jacot described several of
Boeing's efforts in more detail.
The ACESA program was addressing issues associated with precision pointing and
jitter control of a 3-mirror-wide field of view optical system such as ASTREX (Advanced
Space Structures Technology Research Experiments). Two types of beams, a low
frequency and a high frequency, were investigated. Piezoceramics were selected for the
actuators in the high frequency beam (p. D-19): G 1195 type ceramics were used for the
actuators which were embedded in Gr/Ep.
In some IRAD work Boeing is examining the use of piezoceramics for sensors and
actuators (p. D-21). Twenty G 119512 piezoceramic devices 13 were attached to each side of
an aluminum square tube' 4 with an epoxy adhesive. A relaxation behavior was noted over
a relatively short period of time: the beam deflection per volt was less than expected. At
first, this was believed to be due to inadequate curing of the epoxy adhesive. It was later
attributed to local deformation of the tube wall. It was suggested that a bimorph actuator
might help sidestep the surface damage problem. Dean Jacot pointed out, however, that
such actuators probably wouldn't work in most structures.
He identified other problems that Boeing has experienced in working with these
ceramic materials/devices. Cracking has been observed in devices embedded and cured in
composites. Attachment of leads to the devices has been difficult. Initially, round wires
were used which led to cracks in the piezoceramics. Now Boeing uses a flat wire attached
with a silver epoxy, even though the control authority is less. In some cases the full force
of the device cannot be obtained, probably due to incomplete curing of the epoxy. He
emphasized that the load-carrying capability of the ceramic needed to be matched to the
load-carrying capability of the composite as closely as possible.
12 G1 195 materials are similar to the Navy Type I materials, PZT-5A.13 The dimensions were as follows: 0.6 in. x 1.55 in. x 0.01 in. thick. The gap between devices was
0.25 in. (along the length, as near as I can tell).14 The dimensions were as follows: -5 feet long x 3 in. x 3 in. x 0.125 in. thick.
2-4
C. DR. BEN WADA (JPL)
Dr. Ben Wada discussed JPL's interests in piezoelectric actuators for dynamic and
static control such as might be of interest for large truss structures or for attainment of
micron level surface accuracies for large mirrors. Accuracy is the key since loads are low.
JPL's active truss member concept includes a high voltage piezoelcctric motor.1 5
The motor consists of a stack of cylindrical rings of piezoceramic/silver/epoxy/Sn/Cu/Sn/
epoxy/Ag/piezoceramic combinations (p. D-26). Basic characteristics of the current active
member are as follows (p. D-27): a maximum operating voltage of 1000 V with a normal
bias voltage of 500 V; ceramic wafer thicknesses of 1 mm (39.4 mils); displacement of
63.4 gm (2.5 mils) at 1 Hz; a force of 505 N (114 lb) at 1 Hz; hysteresis of 16 percent at 1Hz; and power of 3.77 W for 25 Hz at maximum voltage. The maximum stroke is about
50 gm. In actuality, what JPL thought they were getting from the manufacturer (non-
U.S.) and what they got were not the same. Several of the motors failed and weredisassembled to determine why. Vertical and horizontal cracks as well as voids were
observed in the stacks. It was not clear if these flaws existed before the device was tested,i.e., if they were manufacturing defects. The wafers were specified to be 1 mm thick but
measurements showed that they were 0.5 mm thick. The specific composition was also
unknown.
An alternate design that uses piezoceramic discs without holes has also beeninvestigated. 16 The piezoelectric stack consists of 66 wafers, each 16 mm diameter x
0.8 mm thick. The maximum stroke at 150 V is 45 p.m and the maximum load is -200 lb.The displacement per volt of the in-line actuators was found to depend on frequency. It is
not yet known if the frequency dependency is related to the specific material or is a result of
the structural dynamics. The actuators also elongate over time: peak-to-peak displacement
increases over months (p. D-33).
Piezoelectric strip actuators are used in deformable composite reflectors 17 for staticcontrol, the requirements for which are more stringent than for dynamic control. The PZT-5H actuator strips are 3 in. x 0.5 in. x 0.039 in. thick. The strain level is 2 gm/in, at
500 V at an operating temperature of 200 K; both tensile and compressive strains are
15 The dimensions of the motor are 70 mm long x 15 mm outer diameter, 7 mm inner diameter.16 See the attached paper, page D.41, for more detailed information on the reflector support structure using
the in-line actuator.17 The radius of curvature is 7.6 meters.
2-5
important. Space environmental factors such as radiation, atomic oxygen, and temperature
as well as operational loads were mentioned as potential concerns. Dr. Wada indicated thatJPL would like to be able to maintain the desired shape by applying a voltage periodically
rather than continuously. What they are looking for is a hysteretic actuator. Some generallifetimes of these devices in space were predicted using a safety factor of four (p. D-37):for a random loading cycle, 0 to 50 0C, the device should be able to withstand 360,000cycles; for a sine wave loading cycle, same temperatures, -135,000 cycles are required.For a ground calibration test, using a compressive preloaded condition and low vibrationlevels at 100 K to 323 K, 20 million cycles are necessary. For space operation using thesame vibration levels and a preloaded condition, the number of cycles for Mode 1 is1.2 million, for Mode 2, 61,000. Apparently these devices have not been particularly
reliable in application so far.
Dr. Wada identified a number of technical areas that need to be addressed for both
the dynamic and static control devices. Issues for both include understanding fracturemechanics, strength properties and load-carrying capabilities, temperature limitations,flexibility in bending, bonds and interfaces, reliability, and heat output. Additional issuesassociated with static control include creep, stability, frequency, hysteresis, aging, and loadeffects on the piezoceramic/interface bonds; stroke capabilities;' 8 application of continuous •power, and d13 coupling.
Dr. Wada concluded by stating that there was value added, though this value wasnot quantified: performance goals could be met; reliance on analysis could be less;reliability of the system would be improved; overall system design and possibly ground 0validation tests could be simplified. Whether or not the use of these actuators will be cost
effective remains to be seen.
D. DR. JOHN BREAKWELL (LOCKHEED)
Dr. Breakwell discussed the Space Integrated Controls Experiment program(SPICE). SPICE is a precision optical structure consisting of the following: a large,
segmented primary mirror, a 250-strut bulkhead truss structure supporting the mirror, and atripod secondary mirror support system (p. D-51). This experiment is directed toward 9
18 JPL is interested in actuators with longer strokes from 45 to 1000 pim or inchworm-type actuators that 0are <I in. diameter and 3 to 6 in. long.
2-6
0
addressing line-of-sight pointing stability for the SDIO directed energy weapon systems ofthe future.
Requirements (p. D-54) for the proof mass actuators currently being developed/used for SPICE include a peak force capability of 60 N, a peak-to-peak stroke-proof massproduct of 8.2 kg-cm, power of <30 W, and a frequency range of up to 500 Hz.
The V-struts being used for SPICE contain viscoelastic damping material. Thesestruts have been tested at low frequency (<55 Hz) to determine complex stiffness as acontinuous function of frequency. Control system performance has also been evaluated. Aglobal high authority control/low authority control/passive damping combination is believed
to provide the best performance: attenuation of 100 to 1 can be achieved (p. D-57).
In conclusion, Dr. Breakwell indicated that precise design and modelling of theactuator devices were needed. Determination of interactions of these devices with otherparts of the structure/system will also be critical. He stated that the major point was thatthere is an established procedure in order for a prime contractor to use these devices.
E. MR. RICHARD ALEXIUS (MARTIN MARIETTA-ORLANDO)
The Fast Acting Control Thruster program (FACT) was presented by Mr. RichardAlexius. The overall objective of this effort is to analyze, design, fabricate, and test alightweight, ultrafast, linearly proportional thruster for various Army/SDI applications.The device, illustrated on page D-61, contains an electrostrictive 19 material stack. An
elastomer is used with large and small pistons to amplify the motion (p. D-63): the outputmotion is directly proportional to the piston area ratio while the output force is inversely
proportional to it. The stack has a 5-to- I length over diameter ratio. Tests show (p. D-68)that with a 50-lb input load and an input-to-output amplification of 1-to-10, the output loadis 5 lb. A 96 percent motion conversion is obtained but the stack does not return to its
original position: 0.0025 in. are lost after the first stroke. Apparently there is a friction
problem with the motion amplifier and a compliance problem with the ceramic stack.20 Thecompliance problem is being addressed via lapping stack ends so that they are smooth, flat,and perpendicular to the thruster bore mold line. Larger diameter stacks are also beinginvestigated to reduce the effect of bending. Ben Wada indicated that JPL utilizedmechanical fixtures at the ends of their actuator to avoid a similar bending problem.
19 The elecrosictor material that has been tested in the device is 100AVX 6-8.20 Eight out of 10 stacks have failed due to stack bending.
2-7
When questioned about the processing of the ceramics, Mr. Alexius stated that theywere cold-isostatically pressed to 30 ksi before firing at 1200 'C; the wafers were not hot-isostatically pressed after firing. The electrodes run along the sides of the actuator stack. 0
Even though cracks have been observed at the tabs the stress concentration is believed to be
low. According to Mr. Alexius, the disbond between the electrode and the ceramic is alimiting factor. The actuators have been cycled up to 0.5 billion cycles to failure without
load but have not yet been fatigue tested under load. 0
F. DR. ROBERT GLASER (JPL)
JPL is examining the application of piezoceramics for cryocooler vibration isolationapplications. The flight configuration illustrated on page D-73 is the actual size of the
cryocooler to be flown. More typical piezoceramic actuators are being evaluated to dampthe rest of the cryocooler, to prevent the cold finger from moving. These actuators
(p. D-74) are of the Physik P-842.10 Low Voltage Piezo Translator type,21 about 1.5 in.long. Specific capabilities of the devices include an operating voltage of-20 V to +120 V;
expansions from -3 gtm to 18 pxm; a compressive preload of -67 lb; operating temperature-20 'C to +80 'C; maximum pulling and pushing forces of -67 lb and 180 lb, respectively;
a stiffness of 314 ksi; and a capacitance of 1.8 iF.22 The preload conditions are critical forlaunch. It is believed that low temperature operations should be limited; keep in mind thatthe particular spacecraft conditions are -50 0C to +50 OC. Capacitance is a key property for
power requirements. Dr. Glaser mentioned that these actuators are always one-sided and
need to be centered.
Piezoceramics in the form of an applique are being considered for the cold fingerthe ceramic wraps around the cold finger; its action is out of phase with the movement of
the cold finger. The Vernitron PZT-5H material (p. D-75) is being used as the applique forhigh d31 motion. One of the objectives is to minimize the height as well as thickness of the
ceramic applique/sleeve. Minimization of power requirements to 1 W is also desired.
Some optimization studies were run; results showing trade-offs among height, thickness,
power,2 3 and voltage can be found on page D-76. For lower values of voltage (<200 V)and power (<0.2 W) the height is larger (1.5 in.). For higher values of voltage (>400 V)
and power (>0.5 W) the height is low (<0.8 in.). The applique material apparently exhibits
21 Physik is the name of the company that produces the devices.
22 JPL is using 0 to +28 V, 0grn to 4.2 pun expansion, zero lb pulling force; 52 lb pushing force.
23 The power is observed to increase almost linearly up to the limit of I W.
2-8
marginal behavior at the temperatures at which it is to be used. The ceramic is also brittle
and difficult to handle. Dr. Glaser indicated that the diamond saw is probably causing
cracks in the ceramic.
He also discussed the launch environment that these devices must withstand
(p. D-77). The three worst cases are maximum dynamic pressures, before thrust
termination, and during thrust tail-off. The longitudinal g-forces seem to be much more
severe than lateral g-forces for these conditions. Two tests are of interest for simulatingdynamic loading conditions: sinusoidal vibration tests and random vibration tests. The
sinusoidal vibration test is not critical for this particular application since it is outside thelimits.24 The random vibration test, on the other hand, is critical; it covers the frequency
range from 30 to 2000 Hz.
To conclude, Dr. Glaser stated that lead time from Vemitron for the materials was aproblem. This difficulty has been attributed mainly to changing of designs by JPL after theorder is placed. He believed JPL would buy more ceramics from Vernitron if delivery wasmore prompt. The difficulty of performing development work on the actual materials wasmentioned as a problem. He also indicated that JPL desired larger preloads on the actuators
and the inclusion of flexures with the actuators. Low power flight power supplies wereidentified as another area of concern.
G. DR. VIJAY VARADAN (PENN STATE)
The final speaker in this session was Dr. Vijay Varadan from Penn State. Hediscussed some processing techniques for making these devices and embedding them in
composites. The process illustrated on page D-80 25 is a semiautomated manufacturingsystem concept. High purity ceramic powders can be produced by a sol-gel process orconventional powders can be calcined using a microwave process. This calcination process
results in the formation of small, fine, uniform particles. Powders 26 are characterized bygrain size and purity. Organics are added to the powder (produced by either method) and
ball-milled for 24 hours; viscosity is measured during this step. The ceramic wafers areproduced via a tape casting process. For process control purposes, tape thickness and
dielectric properties are measured, nonintrusively. The tape can then be cut into blanks,
24 The system is designed to be stiff at 300 Hz; the maximum frequency in the sinusoidal test is 100 Hz.Note that the cryocooler itself is driven at 60 Hz.
25 Additional descriptions of the processes can be found on pages D-81 to D-83.26 The piezoceramic material of interest is PZT-5H.
2-9
metallized (for electrodes) 27, and hot-pressed. The final sequence is device fabrication and
firing and sintering. The organic binders are removed during firing; sintering can be done
via the conventional process or by a microwave process. Microwave sintering reduced the
time for sintering from days to about 45 minutes. In an attempt to better understand how to
control the process Dr. Varadan and his students have measured the dielectric constant of
the ceramic in the green state and the fired state; at this time the relationship between the
two values appears to be somewhat tenuous.
The piezoelectric chiral materials developed at Penn State were briefly mentioned.
These materials are capable of larger displacements than conventional materials.
Dr. Varadan indicated that processing is very critical for obtaining the desired properties.
The chiral elements are electroded, though if they are all the same individual elements will
not be electroded. These materials are useful for compliant surfaces or "smart skins."
Microwave processing can also be used to cure the composite in which the actuators
are embedded (pp. D-84 to D-88). The microwave has no effect on the behavior of the 0
actuator: heating of the ceramic is avoided by focussing power at the interface and applying
pressure, as in an autoclave. Microwave processing is also suitable for repair of composite
structures.
Dr. Varadan has developed a portable applicator microwave processing system for
large area joining as well as a taper applicator for strip joining. The applicator is of a
travelling wave type that bonds each layer separately. Bonds made by microwave
processing have been tested in bending and appear to be very strong. The crack is initially
at the joint but grows away from it on further bending.
27 The electrode material is platinum.
2-10
3. PIEZOELECTRIC MATERIALS RESEARCHERSAND ACTUATOR EXPERTS
Five major topical areas were addressed by this group:
1. Currently available materials and performance trade-offs, both for production-qualified materials and limited quantity pilot-scale materials;
2. Developmental materials with promising characteristics;
3. Proposed or predicted actuator materials performance parameters and thecorresponding rationale;
4. Proposed and available multiphase actuator configurations, composites, andmotion amplifiers; and
5. Expected failure mechanisms of actuator materials and potential methods tomaximize and estimate their reliability.
Briefing charts used by each of the speakers in this session can be found in Appendix E.
A. DR. ERIC CROSS (PENN STATE)
Dr. Cross began by describing several types of actuators such as shape memoryalloys (SMAs), piezoelectrics, electrostrictors, contractile polymers, and muscles (p. E-2).The nature of the process by which actuation occurs and the specific phenomena associatedwith that process were also identified. For example, in the piezoelectric ceramics the natureof the process is electro-mechanical transduction; the phenomenon utilized is ferroelectricityin a poled ceramic or in a polymer. Achievable strain levels vary widely: the contractilepolymers and muscles can achieve strain levels on the order of 0.5, with SMAs at 0.1, andpiezoelectrics and electrostrictors at 0.002.
The relationship of piezoelectricity and electrostriction to the strain tensor werediscussed to help focus on the achievement of high strain actuators. The polarization-related piezoelectric constants, bmij, and the polarization-related electrostrictive constants,
Qmnij, are reasonably constant over a range of temperatures. Dr. Cross stated, however,that there is no known solid with high enough values of b or Q and breakdown to allowinduction by an external field of electric polarization values adequate to generate 1 percent
3-1
strain. He indicated that values of spontaneous polarizations, on the other hand, can be
large enough to generate spontaneous strains up to 15 percent in ferroelectric crystals.
There are two options for control of spontaneous strain: (1) phase changes from a non-
polar to a ferroelectric state, and (2) domain changes in ferroelectric crystals that reorient
the spontaneous strain. Dr. Cross also indicated that, only if the electric polarization terms
(PmPn) are homogeneous through the solid, large values of strain can theoretically be
induced at zero stress without breaking it. 0
Various mechanisms for changing polarization to control electrostrictive strain were
discussed in order of increasing induced polarization:
1. Homogeneous polarization of a paraelectric phase--These materials exhibit"dreadful" performance due to extreme temperature sensitivity; the zero strainstate is stable.
2. Induced polarization in a relaxor ferroelectric (conventional electrostrictor)--These materials do not exhibit drift or aging behavior but are sensitive totemperature; the zero strain state is stable. •
3. Induced polarization change in a poled ceramic (conventional piezoelectric)--The zero strain state is unstable due to aging changes of one of the polarizationstates.
4. Micro- to macrodomain poled relaxer--These materials exhibit hysteresis andare sensitive to temperature; the zero strain state is stable.
5. Field forced phase change--These materials exhibit hysteresis and are sensitiveto temperature; the zero strain state can be stable or unstable depending on thesystem.
6. Field forced domain change--These materials exhibit hysteresis; the zero strainstate is unstable.
Most of these materials can be operated using bias voltage to induce a larger d3 1. There are
problems with the bias field, however.
Dr. Cross next discussed the antiferroelectric to ferroelectric switching in the lead-
lanthanum-zirconate-titanate (PLZT) materials. The polarization-applied field responses are
shown on p. E-8 for compositions identified on page E-7. For these materials linearity of
response is sacrificed for magnitude of strain: strain levels on the order of 0.5 percent can
be achieved. In addition, switching between the two states can occur rapidly, <1 11sec if
the applied field is large, on the order of kV/cm (pp. E- II-E-12). There are however, a
number of factors that affect the performance of these materials: specific composition,
3-2
ambient temperature surface preparation, grain size, porosity, and electrodes. The figure
on page E-13 illustrates the importance of surface preparation: the samples with rough
surfaces show a significant degradation in polarization with an increasing number of
switching cycles; the degradation in polarization for the polished samples is noticeably less.
According to Dr. Cross, up to 109 cycles of fatigue can be achieved with these materials if
one is careful in their preparation.
Microdomain to macrodomain switching was the final subject for discussion in this
presentation (pp. E- 14-E-23). This happens in spin glass materials that are transparent and
capable of achieving large polarizations though they are sensitive to temperature (p. E- 16).
The material is fine grained with grain sizes on the order of microns. 1 The switching is
essentially a volume fraction effect: the polarization is changed by controlling the volume
fraction of microdomains being changed, which also controls the proportional strains. It is
possible to achieve -0.5 percent strain. A deliberate hysteretic response, such as would be
desirable for static control applications, is also achievable. The same factors that affect
domain reversal also apply to microdomain/macrodomain switching. For example
(p. E-21), a conventionally sintered PLZT exhibits a 40 percent decrease in normalized
polarization with an increasing number of cycles while a hot-pressed PLZT exhibits almost
no change. As another example (p. E-20), an 80 percent degradation is observed in hot-
pressed PLZT with a conventionally prepared surface compared to a clean surface.
Dr. Cross indicated that stress concentrations in the ceramics at the electrodes are a factor
and can be reduced with careful preparation.
B. DR. STEPHEN WINZER (MARTIN MARIETTA)
Dr. Winzer discussed the advantages and disadvantages of electrostrictive materials
and multilayer actuators. The performance of the electrostrictors are significantly
dependent on temperature, 2 field, and frequency; the phase transition and capacitive
properties are also issues (p. E-26). These materials do, however, exhibit high energy
densities; large force, high modulus and high strain (up to 0.2 percent) capabilities; low
losses, fast response times; and low hysteresis. Issues associated with the multilayer
actuators include the size, weight, and power requirements for the drive electronics;
I The scale for polarization is on the order of nanometers.2 It was mentioned that, in the PMN-type materials (lead-magnesium-niobate), for a temperature range of
40 °C, a 20 percent drop in strain capability is observed. For other, unidentified materials a smallertemperature range results in a 100 percent drop in strain capability.
3-3
fabrication; cost; reliability; fatigue; accessibility; coupling method (attach or embed); and
compatibility. Dr. Winzer also suggested that it would be desirable to vary the thickness of
the individual actuators in a stack. The multilayer devices are, however, flexible in design 0
and can operate over a wide range of drive voltages.
Martin Marietta has been involved in the development of the lead-magnesium-niobate (PMN) type electrostrictive materials. A total of 75 compositions of PMN with
lead-titanate (PT) and barium-titanate (BT) have been examined.3 Both weak-field (Tmax)
and high-field (strain) behavior are predictable. In terms of weak-field behavior, Tmaxincreases linearly with log frequency; increasing the PT content results in an increase in theslope of the frequency dependence (p. E-28). 9
Actuator fabricat*mn techniques were also described by Dr. Winzer. The approach
is similar to that used by commercial manufacturers of multilayer devices. The actuatorlayers are 20-25 gim thick; electrode materials are Pt and Ag, 4 a stripline electrode runsalong the side of the stack with internal electrodes between the layers; operating voltages 0are -25 V; and the field in the ceramic is 1 MV/m. Conventional ceramic powder and
tapecasting processes are used to make the individual wafers. The actuators are then cut,
electroded, layered, and co-fired. One of the workshop attendees asked if it would be
possible to make a multilayer actuator then cut it so that the d33 poles are the poles that 0
provide the actuation. Dr. Winzer indicated that it might be possible as long as the pieces
held together during the cutting step: adhesion between the electrode and ceramic was
believed to be the difficulty. Dr. Winzer worked on the development of the FACT actuator.
He concluded by stating that electrostrictive materials and actuators were applicable 0
to space systems. Issues still to be addressed include the temperature dependence; field,
and frequency dependence of Tc; fatigue analysis and identification of failure modes in both
materials and devices; and miniaturization of the actuator drive electronics.
C. DR. GENE HAERTLING (CLEMSON)
Dr. Haertling provided an excellent overview of the limitations of piezoelectric and
electrostrictive materials. The charts on pages E-35 to E-38 contain a significant amount ofinformation; only selected highlights will be presented here. Properties and associated
3 These materials can exhibit either an electrostrictive or a piezoelectric response. The piezoelectricrespo e is observed below Tc.
4 There is a difficulty with internal Ag electrodes: the silver can react with lead during firing.
3-4
phenomena for the PLZT ferroelectric ceramics were discussed first Dielectric, optical,
and electrical resistivity properties are singular properties; the others, such as piezoelectric
or electrostrictive properties, are interactive. Singular "henomena include, for example,
photoconductivity (dielectric) and space charge effects (electrical resistivity). Interactive
phenomena include photovoltaic (piezoelectric) or photostrictive (electrostrictive) effects.
A comprehensive list of 18 factors that determine performance of these materials is
provided on page E-36. Included among these are composition, microstructure, aging,
fatigue, and creep. Typical ranges for properties of various materials are shown on
page E-37. Dr. Haertling emphasized the minimization of the tan 8 value. He alsoindicated that the less complex compositions exhibited greater stability in terms of their
properties. Common defects include domain bending wLich affects toughness; water
enhances crack propagation; the combination of humidity and voltage degrades both the
material and the electrodes such that grain separation can occur.
Dr. Haertling also presented some limiting values for selected factors that affectperformance of these materials (p. E-38). He emphasized that the listed values are notthose that would be received from a manufacturer; they are, in fact, more optimistic thoughachievable. For example, the best property reproducibility that can be achieved is thoughtto be +3 percent; a manufacturer might guarantee 10 percent. Note that the electric fieldbreakdown values are for the hot isostatically pressed material.
D. DR. ROBERT NEWNHAM (PENN STATE)
Dr. Robert Newnham first considered the mechanisms by which electromechanicaltransducers operate--piezoelectricity, electrostriction, domain changes, or phase changes--and the associated strain versus electric field response of each (p. E-40). A variety ofcomposite actuator components such as piezoresistors or varistors were briefly described.
A number of examples of composite materials for these devices vere also shown (p. E-42):for instance, rods in a matrix (1-3) or spheres in a matrix (0-3). Work at Penn State fundedby ONR has emphasized connectivity patterns and sensors. The illustration on p. E-43shows how compliance can be controlled by using a sensor and multilayer actuatorcombination: the sensor detects stress and the feedback amplifier reduces the actuatorheight and changes its elastic modulus.
One specific application was highlighted. Toyota uses piezoelectric sensors andactuators to electronically modulate an automobile suspension response (pp. E-44 to E-45):
3-5
a five-layer piezoelectric sensor located on the piston rod of the shock absorber detects road
surface roughness within 2 msec; with a high voltage from the electronic control unit an
88-layer piezoelectric actuator produces a 50-gm displacement on the oil system (5 msec); •
the small displacement is hydraulically amplified to 2 mm, changing the shock absorber
damping force in <16 msec. Several other applications were also mentioned (p. E-46).
Dr. Newnham spent some time comparing and contrasting actuator designs: multi-layer stacks with large forces and small displacements; bimorph benders with small forcesand big displacements; and the moonie disks with intermediate forces and displacements.
The moonie actuators5 consist of a piezoceramic disk sandwiched between two shapedmetal electrodes, each having a flat, moon-shaped cavity at the sandwich/ceramic surface.
Cavity size is a critical parameter (p. E-50): increasing the cavity size increases thehydrostatic piezoelectric charge coefficient, dh; for a given cavity size, increasing pressure
up to about 1000 psi has virtually no effect on dh. The particular choice of metal for the
electrodes is important since the metal does most of the straining during actuation. Thestress state of the metal electrode and/or the ceramic will also be of concern. The bond
between the metal electrode and the ceramic disk is another critical feature. Several types ofbonds including polymers, solder, and solder with metal screens have been investigated.Apparently, these moonie PZT actuators offer a factor of about 5 improvement inperformance/displacement over the multilayer stack actuators (at 1000 V, p. E-51).Moonie actuators fabricated from PMN-PT electrostrictive materials exhibit a factor of
10 improvement in performance over conventional PMN (at 1000 V, p. E-53). The d3 3
coefficient for the moonie actuators is a factor of 10 higher than that of PZT.
Dr. Newnham indicated that PMN moonie actuators stacked to a length of 6 in. have
achieved a 3-mm displacement.
E. MR. FRANK A. TITO (NUWC)
Mr. Tito began his review of high power transducers by discussing limitations of
sona - 'ojectors: media limits such as cavitation; 6 physical limits of the radiated sound thatinfluence the projector size; and internal projector limits such as electrical, mechanical, orthermal factors. The schematic on page E-58 illustrates the relationship of radiated power 0
5 The approximate size is -I cm diameter, 2 to 3 mm thick.6 Cavitation occurs when acoustic pressure exceeds fluid static pressure. The cavitation threshold
increases with increasing depth, frequency (particularly above 10-15 kHz), or shorter pulses •(< 0.005 sec).
3-6
versus frequency: at low frequency levels the applied field can limit the performance; athigher frequencies, stress is the limiting factor.
A number of devices were discussed in more detail (pp. E-62-E-77): hydroacousticprojectors, moving coil projectors, flexural bender bar projectors, flexural disk projector,
Class IV flextensional projectors, and ring projectors. Common features include relativelylarge sizes and weights and high power requirements. These transducers operate over awide range of frequencies--from a few Hz to over 100 kHz. For example, the flexural
bender bar projector (p. E-69) is 5.75 in. long and weighs 15 lb; its frequency range is1300 to 1800 Hz. The Class IV flextensional projector (p. E-73) has dimensions of24 in. x 11 in. (elliptical) x 34 in. high and weighs 1150 lb; its frequency range is 450 to
600 Hz. A variety of materials have been utilized for these devices including PZT, PVDFpiezoelectric polymers, PMN electrostrictives, and rare earth magneto-strictives. Theenergy density (as measured by 0.5Y33S2max, p. E-78) for the rare earth magnetostrictives
at liquid nitrogen temperatures is 50 kJ/m3 , significantly higher than for the other materials:PVDF materials-- 110 J/m 3, PZT4 materials--826 Jim 3, and PMN materials-- 12.6 kJ/m3 .To conclude, Mr. Tito stated that size, reliability, cost, and the power-related source level
were important considerations when examining new materials for these devices.
3-7
4. MATERIALS AND DEVICE MANUFACTURERS
This group was requested to address three major topical areas, as follows:
1. Producibility and repeatability issues in the manufacture of PZT materials andactuators;
2. Availability and associated constraints of processing equipment and testing
facilities as necessary for production-quantity materials and devices; and
3. Estimated costs of setting up such facilities, if necessary.
Briefing charts used by each of the speakers in this session can be found in Appendix F.
0 A. MR. BARRY KOEPKE (ALLIANT TECHSYSTEMS)
Mr. Koepke discussed a possible approach to manufacturing PZT materials withimproved and reproducible electro-mechanical properties, mechanical integrity, and
microstructure. The proposed process would be an adaptation of a chemical precursor0 powder production technique currently used to produce transparent PLZT. Expected
benefits include consistent, reproducible powder production; control of powder chemistry;
homogeneous green density and dopant distribution; and a high quality sintered product.
There are difficulties associated with the typical mixed oxide powder production process
0 illustrated on page F-3: incomplete reaction and heterogeneous phase distribution due to
relatively large starting powders;l inhomogeneous dispersion of dopants; formation of hard
agglomerates due to high calcination temperatures (-900 0C);2 and requirement for hot
pressing to obtain high quality material. 3 Another issue is the binder: binders that burncleanly are desired to prevent contamination; minimal binder concentration is needed for
easier burnout.
The composition, morphology, and particle size distribution of these starting powders are critical.Morphology and size distribution, in particular, affect powder flow and packing. Alliant Techsystemsis interested in using smaller starting powders for faster sintering.
2 Agglomerates can lead to the formation of internal flaws as a result of differential shrinkage.
3 Alliant Techsystems would like to eliminate the hot pressing step by decreasing the powder size via thechemical precursor process. On the other hand, for the tapecasting process a slightly larger particle sizeis desi ed for proper dispersion.
4-1
Time and cost are the main differences between the conventional process and the
proposed chemical precursor process, illustrated on page F-5. With the chemical process
one can obtain a uniform phase distribution; a homogeneous dopant distribution; lower
calcination temperatures leading to smaller agglomerates; 4 and finer particle sizes. The
chemical process may allow for better control of sintered grain size, lower sintering
temperatures, and smaller flaws. The most typical flaws are voids. It was pointed out by amember of the group that while voids can stop cracks they are more likely to initiate them.Flaws are typically examined by inspecting fracture surfaces. Mr. Koepke concluded by
stating that the chemical precursor process appeared to be the best choice for manufacturingactuator ceramics with repeatable and controllable properties, improved mechanicalintegrity, and improved microstructures. He proposed a joint Alliant Techsystems/Sandia
program for optimizing this technology and indicated that the necessary facilities were
already in place.
B. MR. CRAIG NEAR (VERNITRON PIEZOELECTRIC DIVISION) 0
Mr. Near discussed issues associated with development and manufacturing of high
performance PZT materials. Reproducibility and reliability were identified as criticalmaterial performance issues. He began by presenting data on various properties 5 of high-power (hard) and high-strain (soft) PZT materials6 (p. F-9): K, icp, d33, and Qm. Lot-to-
lot variations were estimated to be +/- 5 percent for K, a few percent for Kcp, and
+/- 8 percent for d33; variations for Qm are difficult to determine. The question of aging
was raised. Mr. Near indicated that for the high-power/hard materials the degradation in
properties was less than 5 percent per decade.
Stability of the new hard materials is being examined in terms of applied field,
temperature, pressure, and frequency. Many of these new materials contain multiplecomponents that are very interactive; compositions are, therefore, somewhat complicated. 0
The observed improvements in the performance of these materials have been, so far, only
due to formulation; additional improvements are expected with changes in processing, etc.Other manufacturing issues (p. F- 11) include rheology control in various process steps,
0
4 Mr. Koepke indicated that Alliant Techsystems could increase the strength of these materials by25 percent by eliminating agglomerates.
5 Properties are measured on disks of material.6 The new PZT-SK material is a developmental material with a Tc of about 160 0C. Vernitron is
attempting to obtain a d33 value of 1000 in this material.
4-2
selection of binder materials and ZrO2 sources, 7 and the sintering environment and
temperature control.
Some discussion was generated on multilayer actuator stacks. While Vernitron hasnot really made thin, hard materials for the multilayer stacks, Mr. Near also indicated that,though the domain wall phenomena of the soft materials was not well understood,Vernitron knows how to make multilayer stacks from soft PZT. For example, Vernitron
makes a 12-layer, PZT-5H actuator that operates in the 0 to 400 V range; hysteresis is 7percent at 50 V/mil.8 However, the device must be operated below 1 kHz. Maximumdisplacements are on the order of -4 pim. Mr. Near believed that hard actuator materialsshould be used for space applications since the soft materials will depole under continuousload levels and since the operational frequencies for the soft materials were limited. Multi-
element arrays were mentioned as being useful for generating spatial displacements in
optical applications such as deformable mirrors.
Development issues include not only formulation of compositions, but alsoselection of a chemical processing approach; control of the sintering environment andtemperature; 9 and testing in realistic environments. Costs of powders produced by
different chemical processes vary widely (p. F-13):10 powder produced via hydrothermalsynthesis, costs $8/1b; otner powder costs are $20-30/lb for aqueous co-precipitation,$15/lb for thermal co-precipitation, and $18/lb for sol-gel.11 The hydrothermal processappears to be most attractive in terms of material quality (p. F-14). Chemical processingcan result in improved performance. For example, increases in K, icp, and d33 are noted
for a PZT-5H material derived from special powder processing and special firingprocedures (p. F-16). Variations in measurements of a specific material among vendorswere identified as a problem, specifically. Data for a iPZT-8M material tested using the
DOD- 1 376A high field test approach show significant differences when tested by severalgroups, though all data are within the specification limits.
'7 The quality of the PZT materials is especially sensitive to the Z712 source.8 Vemitron has fabricated 25-layer actuators. In the fabrication of 100-layer actuators parallelism of the
layers is a problem. Vemitron uses a roll compaction process to help eliminate nonuniformthicknesses.
9 Data on p. F-15 indicate the importance of sintering pressure and temperature on various properties ofPZT-4SI and PZT-5K.
10 Particle sizes typically range from 200 A to microns with purity levels ranging from 99.5 percent to99.9 percent.
11 The sol-gel process is considered by Mr. Near to be an R&D process.
4-3
Estimated costs for new facilities were also provided: a hydrothermal plant would
cost -$1.7 million; part manufacturing equipment costs are -$1.5 million; special firing
equipment costs are -$250 thousand; testing equipment would cost -$500 thousand.
C. MR. MELVIN MAIN (EDO CORPORATION)
Edo Corporation is capable of producing large numbers of piezoelectric ceramic
devices. New materials development activities include large geometry products, highvolume products, and extensions of performance of Terfenol-D, PZT, and PMN. The
objective of Mr. Main's presentation was to address performance of piezoceramic devices
in space, in particular at cryogenic temperatures.
The DOD-STD-1376(A)SH test geometry was utilized; the procedure was the
American National Standard C83.24-1962. Data were collected from 5 K to 300 K on
several PZT materials including Type I (EC-64), Type II (EC-65), Type III (EC-67 or
-69), and Type V (EC-70). Graphs on pages F-27-F-33 show the relationships betweenvarious properties 12 and temperature. For example, values for d33 for all materials tested
converge at -50 K while values for d31 appear to converge at -150 K. Results showing
the percent change in a property compared to that at ambient conditions (295 K) are
summarized in tables on pages F-34 and F-35. As an example, the percent change in d33 at100 K is -34 percent, at 5 K, -61 percent.
Mr. Main stated that DOD-STD ceramics would not produce high strains at space
temperatures and would perform poorly relative to their performance under ambient
conditions. The devices would then operate over narrower bandwidths, at lower strains,
and with lower efficiencies. He did not believe that changing the ZrO2:TiO2 ratio of the
PZT materials would result in dramatic improvements in their performance at cryogenic
temperatures. Alternative materials warranting development were PLZT and strontium
titanate.
D. DR. JOHN GALVANI (AVX CERAMICS)
AVX Ceramics is a manufacturer of both multilayer capacitors and actuators.
Dr. Galvani compared the capacitor and actuator manufacturing processes (p. F-38).
AVX's capacitor production capacity is 15 million/day with costs for each on the order of
pennies; actuator capacity, however, is less, on the order of hundreds/day with costs in the
12 Properties include d33, d31 , lcp, I33, KC31, Qm, and K3 3T.
4-4
range of tens of dollars. Devices are presently fabricated according to specificationsprovided by prime contractors. Actuator materials are purchased from Vemitron, ChannelIndustries, or others. 13 Actuator electrode materials are platinum. Actuators, designed tooperate at 150 V, contain as many as 500 layers, 125 of which are active. Limitationsinclude the minimum thickness of the ceramic that can be obtained and firing shrinkage(p. F-39) with respect to the electrode. Bum-out times for the actuators are quite long,60 to 170 hours, compared to those for the capacitors, 0 to 4 hours.
Electrode adhesion/bond strength influences the performance of the multilayeractuators. Dr. Galvani mentioned several different electroding techniques (pp. F-40 toF-41). To prevent the occurence of unpoled, highly stressed regions, electrodes are oftenapplied over the full ceramic surfaces. External "pick-up" paths are then insulated fromopposite polarity edges by a Japanese-developed electrophoretic deposition technique usinga glass/epoxy material. Alternatively, a "loose pack" approach uses a porous ceramic andis designed to relieve these stresses: the porous ceramic is located between the ceramiclayers some distance in from the edge. The diffuse electrode is a functionally gradientmaterial: ceramic particles are deposited with Pt such that the volume fraction of ceramicparticles varies across the cross-section; this requires multiple printing, usually 3 to 5times.
E. MR. LESLIE BOWEN (MATERIALS SYSTEMS)
Materials Systems, Inc., manufactures piezoelectric composites and compositeactuators as well as actuator assemblies. The company is also investigating alternativeceramic processing technologies such as injection molding and device prototyping. Theinjection molding process (p. F-44) is attractive since it allows complex shapes to be easily
fabricated: PZT powders are ball-milled, mixed with an organic binder, and placed in amold; burnout 14 in a controlled atmosphere, air flash, and sintering follow.
Mr. Bowen observed that there were several approaches for enhancing the
displacement performance of piezoelectric actuators: high-strain materials; strainamplification via compound actuator designs; and multiple actuator assemblies. Of thesethe compound actuator assemblies were thought to be most likely for success, though high-strain materials and device designs were also claimed to be in the development stage.
13 Actuat materials are PZT and PMN.14 This step rmoves carbon and binder residues.
4-5
He felt the actuators needed to be designed with respect to system goals, manufacturability,
and cost.
Examples of compound actuator assemblies include piezoceramic/polymer
composites and flextensional strain amplifiers (moonie actuators). Potential issues with the
composites include the development of high-strain piezoceramics, processing of
ceramic/metal/polymer interfaces, and fabrication of complex shapes. Alignment, joining,
packaging, and testing will be important during assembly. The 1-3 piezoceramic/polymer
composites consist of PZT rods embedded in a polymer, perpendicular to the electrode
surface (p. F-47). Issues associated with the use of these composite materials include
handling of large quantities of PZT rods or fibers and assembly of the composites; cost and
application and integration into a real system. Injection molding has been selected to form
the ceramics to net-shape. The ceramic preform would look something like a comb
(p. F-48).
There are a few issues that need to be resolved for the flextensional strain amplifiers 0as well (p. F-49). Included among these are alignment and reproducibility of alignment
during assembly; joining, particularly with reference to the device life; design and reliability
of multiple actuator assemblies; and cost. Currently, Materials Systems is using manual
assembly for pilot-scale quantities of prototype devices. According to Mr. Bowen, the 4
devices would be redesigned for full-scale production to simplify joining and alignment.
4-6
5. DISCUSSION
The purpose of the discussion was to explore potential near-term improvements inPZT actuator performance. A number of research areas providing opportunities for such
improvements were identified throughout the meeting. These are listed as follows: d31 vs.d33 performance; displacement/volt; power requirements; fatigue life at high strains; tensile
strain-to-failure; aging behavior and hysteresis; temperature/environmental effects; ceramic
material/actuator fabrication; surface finish; dimensions and available shapes; materialhandling; electrode materials and their application to the actuator, post-composite fabricationactuator performance; and actuator-structure interfaces. Due to the difficulty of addressingall of these items in detail, the discussion was focused somewhat using several questions as
starting points for additional comments. I
A. MATERIAL IMPROVEMENTS
Lt.Col. Obal inquired if the available PZT materials could be improved. Hebelieved that the community was coming up against theoretical limits in some cases.Dr. Newnham indicated that this was probably true for the hard PZT materials though itmay not apply to the soft materials. Dr. Fred Kahn then asked what was specifically meantby improvement. The general consensus appeared to be that strain performance of the hardmaterials needed to be improved. Dr. Cross stated that the hardening mechanism of thesematerials is diffusion-controlled: if domain walls can be moved it may be possible toincrease their strain capabilities.
Dr. Galvani indicated that the operating window for these actuators needed betterdefinition: required strain levels and temperature ranges, as well as voltage/powerlimitations. Dr. Bronowicki stated that actuators used in the ACESA composite strutsoperated in the range of +/-400 V. TRW is interested in having 2-mil thick Type II
I actuators that can operate in the +/-50 V range for the small SDI interceptors. He thenremarked that interceptors in the homing stage will be operating at their maximum
I Note that often there were several conversations occurring at the same time. This chapter contains the
best information available from the authors notes.
5-1
performance level for about 100 sec; frequencies will be on the order of hundreds of hertz.2
The operating conditions for SD1 surveillance systems will be quite different. TRW would
like to have actuators that operate at 10-12 V/mil for these systems. One of the workshop
attendees said that force levels rather than ,,train should be the requirement for designing the
actuators. In fact, it will probably be a trade-off between voltage and current.
Dr. Bronowicki also indicated that the d33 mode of operation was preferred to the d31 modecurrently used. The d15 mode may also be of interest since it has the largest coefficient.The ability of the material/device to withstand a large number of cycles and, if cracked, to
continue to operate is important.
When questioned about loading on actuators, all the space system users agreed thatthe devices were used in tension. The manufacturers then indicated that this was not agood idea: for structural reliability these active devices should be kept in compression.
The idea of reinforcing the ceramics with whiskers was discussed. While the
toughness of the material can be doubled from 0.6 to 1.2 MPa m-1 / 2 there is a loss inpiezoelectric properties, particularly in d3 3. In addition, careful processing is required.
The flaw size cannot be controlled: the initial flaw size is larger and, therefore, the materialis weaker from the start. Dr. Tuttle from Sandia indicated that the critical parameterslimiting the performance of these ceramic materials (with and without reinforcement) were
flaw size and fracture toughness.
The particular aging rate of the materials is also of concern. Such informationwould be obtained from qualification testing. According to Dr. Pohanka, aging rates are
included as part of the specifications for Navy transducers. In some cases, the Navy
requirements are very stringent and push the quality of the materials.
B. MANUFACTURING PROCESS IMPROVEMENTS
A number of manufacturing issues were raised during the discussion. Lt.Col. Obal
commented that there was usually a willingness on the part of the buyer of advanced
systems to pay more for parts, etc., if what was needed could be obtained. This is
particularly true for space systems where the value of improvement (to the system) is high.
In terms of material handling most agreed that improvements in current
manufacturing methods as well as development of new technology such as tape casting
2 Compare this to the few hertz expected for the large, floppy space structures in which NASA has aninterest.
5-2
were in progress. It was believed that much more was possible. There are now only a
limited number of people capable of producing PZT material via tapecasting. Dr. Tuttle
also stated that the tapecasting process was still in the development stage. He did think that
a chemical processing approach could significantly improve material yields over the
conventional oxide mixing approach: a conventional oxide mixing approach was said byDr. Tuttle to have a yield of about 10 percent while yields for the chemical approach are
over 95 percent. Such a process would address material reproducibility issues but notreliability. Others would disagree with this assessment. Dr. Tuttle indicated that devicereliability would be expected to increase if chemically processed materials were utilized.However, one of the other attendees mentioned that chemical processing was not necessaryto eliminate aggregates in calcining: it could be done by screening, for instance.3
Several of the attendees indicated that problems are induced in the actuator ceramicsprior to assembly. Surface and subsurface cracks introduced by machining, for example,
can grow, eventually causing catastrophic failure. Some end users mentioned that foractuator layers 5 mils thick planar materials could not be obtained from vendors. Others
felt that this was not a problem.
Mr. Main believed that the limitation was transducer design, not material. Mr. Titostated that material was a problem 8 to 15 years ago. He indicated that vendor data (bookvalues) are used to design the actuators. However, what usually happens is that pieces are
ordered from the vendors, and properties which often do not agree with the book values aremeasured. Apparently, when a large number of layers are used in a device the variability in
properties becomes less noticeable.
The microwave processing technology was thought to be an interesting possibilityfor implementing these devices into composites.
C. TESTING, NONDESTRUCTIVE EVALUATION (NDE)
Reliability of these devices, clearly an issue, is somewhat related to availability.One person mentioned the difficulty of obtaining adequate data on device performance to
determine statistical limits because of limited availability.
The issues associated with qualification testing were discussed briefly. For spacesystems, "shake and bake" tests are required: these are basically vibration and thermalcycling tests. The need to be sure that qualification testing of the devices is not in the
3 This was thought to be a cost issue.
5-3
region where permanent damage occurs was emphasized. Dr. Galvani stated that there isno good way of testing the devices under various loading conditions: AVX tests for voidsand delaminations between the layers; proof tests4 can only be performed at low levels sothat damage is not induced.
Dr. Bronowicki discussed TRW's experience in measuring electrical characteristicsbefore and after loading of actuators: capacitance measurements were within 4 percent ofthe expected value. Values for Ed31 were at least as good as the book value, sometimeseven better. Cracks, however, were observed between the Ni-plated electrodes and thePZT; the PZT itself was cracked. It was pointed out that these cracks can close up duringactuation. According to Dr. Bronowicki, for a system designer to accept this technology,data from a flight experiment such as A E would be necessary. In addition, vibrationtesting in a thermal chamber would also be desirable. He suggested that efforts similar tothose currently in progress in the SDIO M&S program be continued. The utility of astandard procurement specification with appropriate inspection techniques was highlighted.
0
4 Some of the devices arm tested in tension.
5-4
APPENDIX A
ADAPTIVE STRUCTURES PROGRAMS FOR THE
STRATEGIC DEFENSE INITIATIVE
ORGANIZATION
ADAPTIVE STRUCTURES PROGRAMS FOR THESTRATEGIC DEFENSE INITIATIVE ORGANIZATION
LLCol. MKchael Oba"The Pentagon
Washington, DC
Dr. Janet M. Sater"Institute for Defense Analyses
Alexandria. Virginia
AbtZract
In thn currently envisioned architecture none of the Several different concepts for "Adaptive Structures"Strategic Defense System (SDS) elements to be deployed may be found in the open literature and in the structureswill receive scheduled maintenance. Assessments of community at large 1-6. for ezample. The Strategic Defenseperformance capability due to changes caused by the Initiative Organization Materials and Structures (SDIO M&S)uncertain effects of environments will be difficult, at best. Program has proposed an alternative concept illustrated inIn addition, the system will have limited ability to adjust in Figure 1. Different types of sensors, either embedded in ororder to maintain its required perforn, ance levels. The attached to certain structures, are used to measure specificMaterials and Structures Office of the Strategic Defense environmental features and to perform various subsystemInitiative Organization (SDIO) has begun to address diagnostics. These real-time passive devices can providesolutions to these potential difficulties via an adaptive several functions such as structural health monitoring forstructures technology program that combines health and identification, status, and propagation of cracks. threatenvironment monitoring with static and dynamic structural detection measurements: natural environment measurementscontrol. Conceivable system benefits include on-orbit including radiation, atomic oxygen, loads, dynamic andsystem health monitoring and reporting, threat attack static states. and thermal states: and monitoring of systemwarning and assessment, and improved target tracking and states. The sensory information obtained fromhit-to-kill performance, measurements and subsystem diagnostics is then orocessed
and can be stored on-site or telemetered to another location.ltrdngt This information can also be utilized for static and dynamic
structural controls via appropriate feedback loops usingThe Strategic Defense Initiative Organization (SDIO) active structures containing actuators. The actuator devices
has undergone a dramatic change in its mission architecture, can be static such as for shape control or dynamic such asThe system envisioned during the early years was concerned for vibration suppression: acoustic and propulsion deviceswith the destruction of a substantial portion of a massive may also be used. Applications of this technology are onlySoviet attack involving thousands of nu.lear re-entry now becoming achievable as a result of developments invehicles. Its purpose was to provide sufficient uncertainty to microprocessors and miniature sensors. ConceivableSoviet forces to enhance deterrence. The goal of the new benefits include on-orbit system health monitoring andarchitecture, Global Protection Against Limited Strikes reporting and threat attack warning and assessment via a(GPALS), is to prevent from one to. perhaps, hundreds of combination of sensory structures and informationattacking missiles from reaching their targeL This processing, and improved target tracking and hit-to-killarchitecture may, however, have difficulty achieving its performance using a combination of sensory and activeperfnrmance goal than the previously designed systems. structures with information processing.The GPALS systems are to be designed to allow no"leakers - no penetration of US/allies air space by attacking Although adaptive structures offer some verymissiles. With the proliferation of nuclear and missile attractive features for these complex. autonomous systemstechnologies the threat against the US and its allies is of there are many issues to be resolved. For sensory structuresparticular concern. questions remain on sensor attachment methods; their
durability in natural and threat environments: the number ofTo attain this goal the elements that constitute the sensors requiredb sensor placement and any associated
GPALS architecture must meet performance and reliability constraints; choice of analog or digital output: general sensorconstraints much beyond those of current military weapon performance: and effects of electromagnetic interactions.systems. The elements must be able to react quickly and Information processing issues include selection of localperform optimally after remaining dormant for an extended and/or global control approaches: development ofperiod of time. Two critical components of the GPALS degradation protocol. and selection of a numericallclassical,architecture consist of space-based autonomous surveillance symbolic/rule-based, or a neural network control theoryand defensive elements. None of these elements will receive approach. Issues for active structures include active materialscheduled maintenance. Assessments of performance performancec: type of active device: methods for energycapability due to changes caused by the uncertain effects of coupling: device durability in natural and threatnatural environmental aging and man-made threats will be environments; device placement and any associateddifficalt, at best. In addition, the system will have limited Constraints: and power requiremelts. And. finally, in orderability to adjust critical structural components in order to for system designers to accrot this technology, it must bemaintain its required performance levels. A solution to thesepotential difficulties is adaptive structures technology thatcombines health and environmental monitoring with threat s Other roeMies of the system might also be controlled by suchattack warning and assessment capabilities and static and feedback loops: thermal Droperies, optical properties.dynamic sructural convol. Wornm properties. e.
b A larle mmber of o sophisucitm omnr a few Yvont semmn
could be malized.Menibe AlAA c This includes. for example, the ability of F rmic maenals to
umhstml high soams am maty Cycles.
A-i
least intrusive to the design in terms of weight, power, and assess local structural damage and to determine the
reliability. These issues are being addressed to some degree remaining structural performance capability. The operational
by the SDIO M&S Program as well as by other researchers, requirements and specific environments need to be defined inorder to select appropriate sensors and devices and locations,
M&S Adasrfive Structures Promerams thereof. Acoustic emission and optical methods are beingexamined. Typical sensors may consist of piezoelectrics,
The M&S program intends to leverage outside capacitance-type sensors, and fiber optics.e A proposedresearch whenever possible. However, in order to provide acoustic emission technique uses a passive plate wavethe most appropriate adaptive structures technologies for the approach that recognizes source orientation effects and theStrategic Defense Systems M&S efforts are currently true nature of the waveform as a function of structuralfocusing on the application of adaptive structures technology geometry. Several types of fiber optic sensor- are also beingto provide SDS space elements: on-orbit system health considered in a two-phase program: in Phase 1, specklemonitoring and reporting and threat attack warning and modulation and multiplexed interferomeoic arrays (shown inassessment via a combination of sensory structures and Figure 3) to detect occurrence of impact and location: and ininformation processing: improved target tracking and hit-to- Phase 2, distributed strain rosettes to detect deformation.kill performance via a combination of sensory and active The sensors, to be located on critical areas of the spacecraft,structures with information processing. A M&S program are also expected to be least intrusive to the system in termschart showing points of contact and areas of research can be of power and weight. This technology is applicable to afound in Table I. The general approach is through ground- broad range of systems; examples are BP and BE opticalbased demonstrations leading to generic structural space sensor, antenna, heat exchanger, and solar array surfaces.experiments as appropriate. Examples of some of theseprograms follow. To actively determine the health of critical optical
coatings and other materials and assess spacecraft self-Sengory Structures contamination in the orbiting environment, M&S is
developing the Space Active Modular Materials ExperimentsFour areas M&S is currently focusing on in this area (SAMMES) program. The SAMMES modules, located on a
are real-time assessments of the state of (I) critical moving generic spacecraft, and the system configuration aremechanical assemblies (MMAs) and momentum transfer illustrated in Figure 4. Though not the original goal of thisdevices; (2) micrometeoroids and debris (MM&D) program, an additional benefit of SAMMES may be theidentification: (3) real-time evaluation of critical materials capability to provide spacecraft with lightweight, low power,deterioration and spacecraft contamination; and (4) threat modular avionics for active health monitoring of missionattack warning and assessment. critical materials. Recent findings from the Long Duration
Exposure Facility (LDEF) suggest that contamination mayMoving mechanical assemblies and momentum pose serious problems for SDIO-like sensor systems.
transfer devices are mission critical components on many SAMMES may also offer the potential for active monitoringDoD space assets: if the device fails the system cannot of contamination.perform its mission. Examples are illustrated in Table 2.Wright Laboratories has initiated a M&S program to develop To field a fleet of surveillance and space-baseda health monitoring system for MMAs to help address interceptors or high value directed energy weapon (DEW)potential failure problems. An example of a smart platforms will require continuous knowledge of the threattribomechanism is illustrated in Figure 2. Major segments of environment. Potential kill mechanisms range from thethe program include the following: (1) identification of subtle (i.e., damage to parts of an electronic system fromvibration and torque signal signaturesd for bearing structural microwaves) to the direct (i.e., major structural damage frommechanism and lubricant failures: (2) use of embedded or a nuclear explosion or a high-power laser). The externalattached sensors to identify changes in acoustic or thermal surface of the space asset provides an opportunity to embedsignatures of the device; (3) development of an on-board or attach sensors that can identify and measure the threat.control system to enable corrective action such as activation An additional benefit will be the ability to distinguishof adaptable bearing preload or of exercise protocol for between system failures caused by natural or man-madefretting suppression: and (4) analyses of actual and predicted environments. The M&S Program has initiated a program tosignatures to determine the expected remaining life of the develop a "Smart Skin": a prototype. light weight. lowdevice. This technology, applicable to any space assets power sensor skin. These sevlsors will be able to measurehaving MMAs. would provide on-orbit capabilities to control radio frequency (RF), laser, and nuclear energy depositionor alter performance to extend system life. on the satellite. These sensors include conformal antennae
for RF detection, coated pyroelectric films for laser detertionOne of SDIO's concerns for SDS elements is the (Figure 5), and fiber optics for nuclear detection. The
degrading effect of MM&D or kinetic energy weapons sensors and initial signal conditioning are embedded or(KEW) attack over t, "te on critical sensor structures. To attached to the spacecraft skin which provides thermal andprovide real-time as. ;ssment of these impacts, NRL is space environment protection for the devices. Informationchartered to investigate technologies for impact detection, from the skin is provided to on-orbit or ground-basedlocation identification, and damage assessment sensor systems that would identify the source of the energy and itssystems. These sensor systems will be demonstrated via damage capability. Since some of these sensors do requireextensive ground and/or flight testing. M&S is leveragingongoing work for DARPA on application of such sensorsystems in submarine hulls for damage detection and Fiber optics have advantages that include light weight. immunityanalysis. The general approach is to detect, locate, and from electra-magnetic interactions. electrcal passivity, low powet
ConsumPbon, minimal leads in and out. and no direct current drift orbias offset. Disadvantages depend on the specific sensor type andmay include inadequate sensitivity, requirement for integrated
These signature outputs are to be correlated with specific failure measurement over the whole fiber length. complex signal prnce sing.mechanims. Under another NI&S tribology program Lockheed is fabrication of many reflective splices, requirement for multiplecorrelting torque signals with failure mechanisms of super-dense wavelength light sources, temperature sensitivity, and the need forMoS2-coated bearngs. high speed electonics.
A-2
surface mounting, the skin itself must not interfere with the promoted significant damping and very short settling times.
thermal balance of the spacecraft or require an excessive Good agreement was obtained between control simulations
amount of surface area. A variety of materials have been and experiments. Additional work is still needed, however.
examined for the skin structure. Combinations of advanced to address other issues such as active material development
thermoplastics with shuttle tile ceramics appear promising to obtain better actuator performance/watt, sensor and
for structural rigidity and thermal balance features. The actuator placement, miniaturization of circuits and power
advantage of real-time threat detection and assessment is the conditioning, advanced control theory for adaptable control,
ability to provide critical information for the SDS fleet "user" neural network development for the "intelligent'" aspects, and
to enhance survivability of those elements. This technology survivability in the space and threat environment.
would be applicable to all US and Allied space assets.Forward, Swigert, and Obal9 completed one of the
.ensorv/Active Strictures first successful demonstrations of the application of surface-SDS elements require extreme tracking capabilities to mounted piezoceramic sensors and actuators for vibration
meet mission performance goals. Sensor jitter from control in a SDIO-like directed energy weapon system - the
structural response seriously degrades this capability. Active AF Airborne Laser Laboratory. A substantial amount ofand passive vibration suppression technologies provide a jitter reduction on a cavity resonator mirror was achievedmethod of adaptable jitter control. Vibration control using with the combination of a passive tuned mass damper and anpiezoelectric devices as sensing and actuating elements has active rate feedback vibration control network that includedbeen in development for at least two decades. BK. Wada's piezoceramic devices. Unfortunately, the state of theoverview 5 of adaptive structures identifies research efforts technology at that time led to excessive high voltage powerundertaken in the early 1980's. This research has been requirements. ACESA demonstrated an order of magnitudesignificantly extended by industry and academia to the point improvement in both actuator coupling and powerwhere integration into fielded systems is more appropriate. requirements.
The ACESA (Advanced Composites with Embedded To meet current SDIO system power constraintsSensors and Actuators) program was an early M&S initiative using active vibration control, the performance of theseinto the sensory/active structures arena. Overall goals for the active materials must be improved in order to obtain greaterprogram included design, fabrication, testing. and evaluation strains for a given input electrical load. To that end a taskof composite components containing embedded sensors, has been initiated at the Naval Research Laboratory (NRL) toactuators, and microprocessors. A number of parameters improve toughness and durability of piezoelectric andimportant in the selection of the sensors and actuators were electrostrictive actuator materials in order to obtair largeridentified: accuracy, dynamic range, frequency response, deflections. possibly up to 10 times that of conventionallinearity, noise rejection, health monitoring capability, materials. The M&S Program is also looking for significantnetworking, device cost. survivability in natural and threat improvements in the number of cycles these materials canenvironments, te:hnology readiness, embedding capability, withstand under high/maximum strain conditions to addressand material processing conditions. A number of sensors reliability concerns. Improvements in materialand actuators, among ',hich are fiber optics for sensors and reproducibility and manufacturing quality are also desired:shape memory alloys (SMAs) and piezoelectric ceramics for i.e., it is difficult to obtain lead zirconate-lead titanate (PZT)both sensors and actuators, were examined in light of the materials having small variances in performance parametersabove parameters. The ability to use relatively simple drive from current vendors. Some interesting work is being doneelectronics combined with other beneficial properties( led to by Litton Optical Systems in conjunction with Thethe selection of piezoelectric ceramics for both sensors and Pennsylvania State University on the development ofactuators. discrete multi-layer actuators made from lead magnesium
niobate doped with lead titanate (PMN:PT) 1O. Results ofSome difficulties with embedding these and other their parametric studies showed a factor of -3 improvement
devices into the composites were noted 7. 8: fiber optic in achievable strain as well as improvements in materialbreakage. piezoelectric ceramic electrical insulation and reproducibility and manufacturing yield. Included in thecrackingg, mandrel thermal expansion, longitudinal mandrel NRL task are specific research studies on methods tobowing, and electrical insulation of shape memory alloy eliminate extrinsic voltage breakdown; the use ofdevices. Additional issues included mechanics of the compositional chemistry, i.e., doped barium-containingdevice/structure interface during compression or tension, PZT, to increase strain performance: and development ofminiaturization for weight and power management, and composite piezoelectric and electrostrictive materialscentral vs. distributed analog-to-digital (A/D) processing, containing whiskers or fibers to improve strength andGenerally, though, characterization tests verified that the toughness and to obtain higher c/a ratios. Methods forsensitivities and dynamic ranges of the sensors and actuators integrating these piezoelectric devicesh into different actuatormet the requirements derived from the system studies. The design configurations are also being investigated. Improvedactive structurAl control tests showed that the devices actuator materials would be applicable over a broad range of
structural systems to provide better energy coupling.
Recognizing that the option of embedded sensors andPiezoceramic sensor materials such as lead zironate-lead tiutmate actuators will not be permitted for all structures, i a concept(PZT) exhibit low temperature and radiation sensitivity and highstrain sensitivity. As sensors they have a large dynamic range and afrequency range exceeding the kilz level. In trms of actuation PZTdevices have a quick response time coupled with high efficiency and a The JaPamese we transitioning advanced piezoelectric devices into apotentially large force authorty. number of commercial products. i.e.. autofocus mechanisms for
One solution to thi difficulty was to encapsulate die PZT deice cameras, curtain pullers. aerators for fish tanks, laser printer heads,with fiberglass or kevlar/epoxy prior to placement within the
structure. This step is particularly crucial if the device is to be For eximple senors and actuator may not be able to endure severeembedded into a graphite-reinforced composite since the fibers am composite component manufacturing processes. Or. vibrationconductive and will ground the high voltage Circums necessary to problems may not have been anticipated or identified during thepower the device. design process thus requiring a retofitned solution.
IA A-3
for a space-durable modular patch is being developed. The mounting tripod (Figure 7) with numerous embedded PZT
patch integrates the sensor, actuator, controller, and power sensors and actuators and corresponding control avionics
conditioning electronics into a single package that can be will perform various vibration suppression and adaptive
bonded to the structure and interfaced directly with the control experiments. Issues being examined include
power system. An individual patch could exhibit local mechanics of piezoelectric deviceistructure interfaces,vibrational control or could work in concert with other durability of the devices and structural materials in space,patches in a global manner. Naturally, the patch must be system idenification of the structure, methods to change
space-durable and engineered in such a manner as to provide structural stiffness for re-identification, and miniaturizationoptimized coupling between the actuator and the structure. of power and control devices. Three years of on-orbit
Expected benefits include significantly reduced control performance and space effects data are to be obtained via asystem weight and volume and, potentially, increased space flight.damping. This technology could be used to retrofitstructural components for which embedding such devices is An ultra-fast (-I msec), lightweight (-0.1 Ib),not an option: systems that undergo severe composite linearly proportional control thruster (10 lb maximum force)manufacturing process conditions or existing systems having called the Fast Acting Control Thruster (FACT), is beingvibration problems. developed to improve vehicle attitude control for various
SDS elements. The device, 0.625 inches (diameter) by 2.5Such active vibration suppression concepts will be inches (length). consists of a piezoelectric/electrostrictive
demonstrated on the Advanced Materials Application to actuator, an elastomeric motion amplifier, and a cold gas,Space Structures (AMASS) program. This program high force-gain valve. It is illustrated in Figure 8. Aincludes studies on dynamic decoupling of solar array number of issues are being addressed in this program: activesupport structure during rapid motions. Figure 6 is an material hysteresis, general performance, and fatigue limits:illustration of the concept. Solar arrays, utilized on a active material stack design and manufacturing; motionnumber of existing satellites, are expected to be used for amplification device performance: throat positionseveral SDI systems. determination and feedback: and reductions in size, weight
and power requirements of drive electronics.As mentioned previously, sensor jitter continues to
be an issue with the advanced SDS assets. One of the Transfer of an innovative spacecraft technology such 4greatest sources of sensor jitter. given a quiescent spacecraft, as adaptive structures into developing spacecraft is the lastis the cryocooler itselfI 1. 12. A project was initiated by step in the maturation of a technology and can be aM&S with JPL to address this difficult problem. A PZT formidable challenge. Spacecraft designers are justifiablydevice is being used to isolate the motion of a cryocooler cautious in accepting a new technology without flightcold finger on an existing, advanced Stirling cryocooler. heritage. Therefore, M&S has initiated plans for a programJPL expects to demonstrate a factor of 2 to 5 reduction in called TechSat. Its purpose is to provide an on-orbit, multi-transmitted vibratory forces in the 0 to 200 Hz range. Such discipline experiment platform for testing and validatinga cryocooler system is expected to be flown on a UK satellite promising technologies important to SDI space assets. Anin FY93 to obtain space durability data in a high radiation initial list of candidate experiments for TechSat includesenvironment. The technology is applicable to extremely several adaptive structure demonstrations that build onsmall cryocooler systems intended for BP or BE where advances described elsewhere in this paper.cryocooler-induced vibrations couple to the focal plane arraysensor and lead to increased jitter. System advantages obtained from use of adaptivestructures technologies are expected to include increased
To provide adaptive control of the sensor critical agility, pointing precision, stability, and shape/alignmentstructures, the state of the plant must be known at all times. control, all of which contribute to enhanced target trackingThe M&S program is pursuing demonstrations of system and hit-to-kill performance. These technologies would beand parameter identification (ID) technologies to meet this applicable to a number of SDI systems such as BP, BE,requirement. To address this issue researchers at the Jet Ground-Based Interceptor (GBI), E21, THAAD, and NeutralPropulsion Laboratory (JPL) 13 have been examining Particle Beam (NPB).methods for system ID of open and closed loop structuralsystems in space using active members on ground test Sumararticles. Experiments were performed on a cantilever truss A number of potentially significant benefits for SDSstructures. Their reults indicate that active member% can be elements using adaptive structures technologies have beenused as an excitation source for on-orbit system identified: on-orbit system health monitoring and reporting,identification for both closed and open loop systems. The threat attack warning and assessment, improved targetactive members provide a modal excitation source to aid in tracking and hit-to-kill performance. The current M&Sidentification of low amplitude dynamic characteristics that programs were basically selected and initiated during FY91.are important in large, precision space structures. The Communication and coordination with the SDS programdifficulty of simulating on-orbit conditions on the ground for elements (BP, BE, etc.) is a continuing effort. All the M&Ssystem/structural identification is well known. M&S is programs have planned intermediate ground demonstrationscooperating with the Phillips Laboratory in advancing such to assess their progress since many technical issues still needinvestigations via the Inexpensive Flight Experiment to be resolved. Additionally, all of the adaptive structures(INFLEX) program. On-orbit system and parameter ID technologies developed under the M&S program mustexperiments will be carried out on this space test bed in satisfy the necessary space and threat durability requirementsFY95. as well as provide minimum weight, power, and reliability
concerns for the system. Though some elements of adaptiveThe Advanced Control Technology Experiment structures technology are ready for demonstration there are
(ACTEX) will demonstrate many of the adaptive structures issues that remain to be addressed. The SDIO M&Stechnologies being investigated by M&S for improved Program is leveraging existing efforts in and coordinatingtracking and hit-to-kill performance in space. In fact, the with other U.S. government programs and agencies totechnology developed by the ACESA program provides the address some of the most important concerns.basis for this experiment. A small, composite, sensor-
A-4
Ret'eignees
1. "Fiber Optic Smart Skins and Structures." August 1990.Marketing Research Review via NewsNet, based onInternational Resources Development, Report 772.
2. Rogers, C.A., E.S. Chen, and A.F. Findeis. 1988."International Workshop on Intelligent Materials."ONRASIA Scientific lrn'ormation Bulletin, 14 (3): 23-33.
3. Ahmad. lqbal. 1990. "U.S.-Japan Workshop onSmart/intelligent Materials and Systerns."ONRASIAScienific Information Bulletin, 15 (4): 67-75.
4. Brown, A.S. March 1990. "Materials Get Smarter."Aerospace America .30-36.
5. Wada, B.K., J.L. Fanson, and E.F. Crawley. April1990. "Adaptive Structures." Journal of Intelligent MaterialSystems and Structures, 1 (1): 157-174.
6. Wada, B.K. May-June 1990. "Adaptive Structures: AnOverview." Journal of Spacecraft and Rockeu, 27 (3): 330-337.
7. Ikegami, R.., D.G. Wilson, and 1LH. Laakso. June 1990."Advanced Composites With Embedded Sensors andActuators (ACESA)," Boeing Final Report, ContractF04611-88-C-0053, III pages.
8. Bronowicki, A.J., T.L. Mendenhall, and R.M.Manning. April 1990. "Advanced Composites WithEmbedded Sensors and Actuators (ACESA)," TRW FinalReport, Contract F0461 1-88-C-0054, 162 pages.
9. Forward. R.L, CJ. Swigert, and M.Obal. May 1983."Electronic Damping of a large Optical Bench." The Shockand Vibration Bulletin, Parr 4 Damping and MachineryDynanics, Bulletin 53,: 51-61.
10. Wheeler. C.E. and B.G Pazol. 1991. "MultilayerElectrodisplacive Actuators." Ceramic Bulletin, 70 (1): 117-119.
11. Ludwigsen, J.. "Vibration Suppression ofCryogenically Cooled Sensor Systems," white paper,Nichols Research Corporation - Albuquerque, April 25,1991, 5 pages.
12. Ludwigsen, J., "Cryocooler Vibration Effects onSensor Performance," white paper, Nichols ResearchCorporation - Albuquerque, September 18, 1991,5 pages.
13. Kuo, C.P., G-S Chen. P. Pham, and B.K. Wada.April 2.4, 1990. "On-Orbit System Identification UsingActive Members," 31st AIAAIASME/ASCE/AHSIASCStructures. Structural Dynamics and Materials Conference,AIAA-90-1129-CP: 2306-2316.
A-5
Table 1. Materials and Structures Adaptive Strucitores Program Organization FY92
Programs Points Of Contact Phone No.
Sensory Structures F1504 (Air Force)With Information Health Monitoring of Moving Mechanical Mr. Karl Mecklenburg 513-255-2465Processing Assemblies
N1504 (Navy)Sensor Development for Micrometoroid/Debris Dr. Robert Badaliance 202-767-63800
IdentificationS1504 (SD1)
Space Environmental Effects and Contamination LI.Col. Michael Obal 703-693-1663Sensor DevtlopmentsE1504 (Department of Energy)
Sensory Structures Dr. Mark flodgson 505-667-6772
Sensory/Active FIS04 (Air Force)Structures With Passive Damping Application Technology Dr. Alok Das 805-275-5412Information Vibration Suppression for Cryocoolers Mr. Paul Lindquist 513-255-6622Processing Advanced Materials Applications for Space
Stnicttures (AMASS)Advanced Composites with Embedded Sensors
and Actuators (ACESA)Advanced Control Technology Experiment (ACTEX)Modu'lar Control PatchI ligh Frequency Passive Damping Strut DevelopmentOptional PZT Passive DampingAutonomous System IdentificationAdaptive Structural Control
AIS04 (Army)Adaptive Thermal Isolator Mr. Doug Ennis 205-955-1494Fast Acting Control Thruster
N 1504 (Navy)Large Deflection Ceramics for Actuators Dr. Man fred Kahn 202-767-2216
S1504 (SI)IO)System Identification Flight Experiment (INFLEX) Dr. Fred Hadaegh 818-354-8777Techsaz LLCoI. Michael Obal 703-693-13663
Table 2. A History of Tribology Problems in Space and Spacecraft System impacts
Program Tribologicat System Problem ImpactDMS? OLS Sensor Launch Chump Seizur on Launch Pad Single Point Failure. Prohibit
LaunchNAVIGPS Reaction Wheels (4/Satellite) TonlquelTenmure Lubricant Lass and Stavation
Runaway. Pointing ErrorsSkylab CMIG Beaig Failure Prenaure Mission Failure0CDP CMIG Beaig Failur Loss of MissionCDP ATP Harmonic Drive RXcesive Wear Degraded Mission, Possible FailureCDP LArg Ct4Gs (4/Satellite) Runaway Torque; Cage Life Test Failures @ 51/2 Life,
Fracume Lubricant Lubricant Starvation, cageBreakdown During Tat Instabilit&y Using Active Odle
DSCS, CDP Slip Ring Excessive Noise Communication Amennma PointingMission Compromised
Galileo High Gain Antenna Sticking Of Antenna Rib Unfurled Antenna,Pins Due to Dry Lubricant Crippled MissionLOSS
A-6
Measurements Results
I. Enhancedtargettracking and
moniorin
repocin3 . THeathatcE R 0w0amnitngn
S st m eSn eo NA ti ea s s m n
-~ nt S .NoSrcue
Fiur 1. SDI Prreeporeptfindgiv Srct
Rin U SI hratatacY' SinR ETeoeRelTm
Seain waaniablen
N Structures
JjJ\\Failurearing olkttApeunignature
1PreloadFiur 2.Shemuo-nasnThioesmw
ap [prcesso
aoR am
... Distribtited Strain Sensing byWavelength Encoded Gratings
Broadband
Fibr atig Sin.w
FigureS NMermeteorite bp F, Dejection and Damage Aineeit
SAIAMES System Corniguration
ACTNOGW
ACAM
Uaser ArraY. oltages Impact of r m
WN" OdtI rorrym Data psto
ezoYor procesing Sv ,talFesPonse
Surace,4, ul tln Layer
and Slectivecoating
rhrI& nd IMpact
*inner FimtIipect
Siwtre
FigUm S. Schw"'SC ot POWymr Laer Sensor SkIdi.
Soler Allapanel
Actuatos
Fight 6. SclIStl of AMASS SOW MMlE 5%lt"
A-0
Tri-axAccelerometer
BI-axAccelerometer-
Adaptive Tripod
Data Acquisition Sae
ImpedanceSolid Stat HaRecorder
Figure 7. Schemnatlc of Advanced Control Technology Experiment (ACTEX)6-28-91-2M
Inlet eedbac
Inletstomere
144-SMFigureS SchemIcl of FaM Acing Control Thruster.
A-10
APPENDIX B
WORKSHOP AGENDA AND
LIST OF ATTENDEES
Workshop Agenda
Institute for Defense Analyses2001 N. Beauregard Street
Rooms 121-123
7:55 am Welcome J.M. Sater, IDA
8:00 Introduction Lt.Col. M. Obal, SDIO
8:10 Meeting Objectives M. Kahn, NRL
I. Piezoceramic Actuator Application Designers and Users
8:15 ACESA, ACTEX Experiments A. Bronowicki, TRW
8:30 Piezocontrol Benefits D. Jacot, Boeing
8:45 JPL In-Line Truss Actuator B. Wada, JPL
Experience
9:00 DEW Space System Applications J. Breakwell, Lockheed9:15 Fast-Acting Control Thruster R.C. Alexius, Martin Marietta
9:30 Cryocooler Vibration Isolation R. Glaser, JPL9:45 Application of Piezoelectric Materials V. Varadan, Penn State
to Mechanical Response Control
10:00 Break
H. Piezoceranic Materials/Actuator ResearcherExperts
10:15 Practical High Strain Materials E. Cross, Penn State
10:30 Design of Space-Qualified Actuator S. Winzer, Martin Marietta
Materials
10:45 Intrinsic Limitations of High Drive G. Haertling, Clemson
Piezoelectric Materials11:00 Actuator Configurations R. Newnham, Penn State11:15 High Power Transducer Design F. Tito, NUWC
11:30 Actuator Testing and Reliability E. Cross, R. Newnham, Penn State11:45 Composite Materials M. Kahn, NRL
12:00 Lunch
I Speakers, please limit the number of briefing charts for presentation to 5 or 6. Plan for a 10 minutbriefing to allow 5 minutes for questions and discussion. Additional charts may be provided toattendees or used as backup and during the discussion period. You may also bring samples, etc.
B-1
M. Materials/Device Manufacturers:
1:00 pm Bulk Powder Processing - Impacts B. Koepke, Alliant Techsystems
on Reproducibility and Reliability
1:15 Manufacturing Issues with High C. Near, Vermitron Piezoelectric Div.
Performance PZT Materials
1:30 DoD STD 1376 (A) SH Ceramics M. Main, EDO Corp.
Types I, II, I[, V at Cryogenic
Temperatures •
1:45 Manufacturing Constraints of J. Galvani, AVX Ceramics
Co-Fired Multi-Layer Actuators
2:00 Tooling Up to Make Actuator L. Bowen, Materials Systems
Assemblies
2:15 Break
2:30 Discussion and Closing Remarks All
B
0
B-2
List of Attendees
Speakers: J.M. Sater, IDALLCol. M. Obal, SDIO
A. Bronowicki, TRW
D. Jacot, Boeing
B. Wada, JPL
J. Breakwell, Lockheed
R.C. Alexius, Martin Marietta
R. Glaser, JPLV. Varadan, Penn State
E. Cross, Penn State
S. Winzer, Martin Marietta
G. Haertling, ClemsonR. Newnham, Penn State
R. Tito, NUWC
M. Kahn, NRL
B. Koepke, Alliant Techsystems
C. Near, J. Gray, Vernitron Piezoelectric Div. of Morgan Matroc
M. Main, EDO Corp.
J. Galvani, AVX Ceramics
L Bowen, Maerials Systems
Guests: S. Griffin, PL
W. Smith, R. Pohanka, ONR
B. Tuttle, Sandia
G. Homer, NASA-Langley
B-3
APPENDIX C
ADVANCED PIEZOELECTRIC CERAMIC
ACTUATOR MATERIALS FOR
SPACE APPLICATIONS
C')
go-
0go-m
a)* C..)
C Co n 0a) 0/ <o ~ 2~0 c5.0-rC <
az a0 -C C) .'o~~ a L
o =~ 0.*
> c-i
00
-~ 0
~ME Cl)C.,
I. 0
__ a 0DcaC
00c
c) 0 cz 00M..- Ccn
cl) 0 CZcz -0-0
C)~ ~ 0) C L)0) a,
_n ECO~cv CD WW o n
E) EC.,a -a tu 0 0
M 0
a) CC . C
CCm)
C-20
4 %~I~ ~0 C/)
=~ 0. C Cl4,-
2 0 Ec)VPA CZC
C/)Q:
0 c
CZ 0)CZ 0 0
E CL ocOC
0E_ cCZ C:CCD a(/Z 4-"00)D C c EC u-0
Co Oz -0C
0.0 cz
0)CD L.CC CO C0
-C6--a 00Ca~ 0 coC
*z C/-
APPENDIX D
ACESA, ACTEX AND AMASS PZT
MATERIAL NEEDS
7kto t
voM 6
(U
D-1-
00
cc. *2 (U
~ucurP~OW4aU* 1- M
*bd w 05
1b( -4.4 -j5- M 0 - JJ (U
Ovo*", ,uOVA
5-~~~ =~~-* 5
o 0-;o ( a
Vol6 * L
CA-u~'a
>~ W- 0) %. 0 rA V0 A6~C)
= -=. - 5 Q U> d
OVA~ %4 ow OWN I S
* S V 0 0 cuS S
%6.4ajD-:
A~~v DMP9D wgrh~PZT Actuation Comparison
NAVY TYPE I NAVY TYPE
a.0
EN 40
S~0I N2o 20N
0 0
PU04o Ipp
-44£N V.. ~ v.a au~0
PC 0e.. NY Lt4a I..V-61MA ONS..w
-110 -00 *$0 0 Src ioo 40 -8 i , =~l ?.,. ,~ 6.
A45A W -ts.... (v) -IO -eec -1 ; 0o jOe ISo
Av.vej. Vos.YAgg (v)
NAVY TYPE I E. NAVY TYPE r1
30
I 0 N
0,
I0
*1o!
o . .. p0 0
Type I has lower intrinsic hysteresis & better stiffness match to graphite.SS
e IIPhZshig stbe edng rhe/Epy
D)-3
I-l
E 4L
wwo
am
W44 1.4ft
_x CO( , I.
_ N Ionto 'I ~ i
w4, ."q- .
m~z'I NC- Ii..
>4W'm_____
ej~ 0
- : - 4
II II( _ _ w
%-- se - .- - 4 U
%low*wUI
.Olq . p.
- ~ 4eU _ _ __
tu Cmu t
rwU~
4 ,.
hi +1
m U
o~t to h
._. 0W4
mU e-~~
= EEU ".0 0vo
o4 w' go =~ EU
cii
0~ _ EU EUoE0) U U) 4' _ +
0146
EU' C4OqEU4'.~D-5 ~
OF" 4 1 tC 4- 4. 0
N0 aI
> do L IF Oa
0- 44.. . + 'P_ +-r' + We- u
3f )-.- - , ,,,., . .I:to E
4-- 11
OV .0w",
-- / I m- . tu
_ -4 * .0 m Wto = 'm >
00 w v
0.a ' Ii .7/.. o v 2/ - a '
ZU U)4 F.; C4 v0 44v
D-6
damped composites. Two standard Navy PZTcompositions, Types I and 11, were evaluated.The strain limits of the encapsulated wafers inthe lateral 1-direction were tested to
Mechanical Validation of Smart determine the applied strain level at whichStructures point actuation losses occur.
Allen Bronowicki, Robert Betros & Ted Nye " 2.1 COUPON DESCRIPTION: 7.5 mil thickand PZTs were encapsulated in an insulating
Lon McIntyre, Lee Miller & George Dvorskyl medium prior to embeddment in either agraphite/epoxy or a graphite/thermoplastic
TRW Space & Defense (RadeIC - polyarylsulfone) laminate.1 Space Park, R4/1074 The graphite composite lay-up was designed
Redondo Beach, California 90278 to provide a modulus similar to the PZT's (10(310) 813-9124 Msi) being tested. Strain gauges were applied
to the exterior of the completed coupons, andABSTRACT: TRW is developing smart strut the coupon strain was recorded as a function oftechnology for application to spacecraft applied voltage. An example of one of thevibration and shape control on a number of generated plots can be seen in Figure 1.contract and internally funded researchprograms. Performance verification oflightweight composite structures withembedded piezoceramic sensors and actuatorsin the harsh spacecraft environment is a keyrequirement. Graphite epoxy, graphitepolycyanate and graphite thermoplasticmembers have been fabricated with thin lead . *
zirconate-titanate (PZT) actuator and sensor Iwafers embedded in the composite layup. IThese members have then been subjected to . /tension and compression loading, hundreds ofcycles of fatigue loading at levels indicative Tom..
of launch loads, and thermal cycling tests attemperatures found in the hard vacuum ofspace. Results of the testing are promising.
1. INTRODUCTION: Figure I. Initial Actuation Response of NavyTRW has been performing mechanical Type 11 PZT Encapsulated in Thermoset.
validation testing of smart composite membersunder the Advanced Composites withEmbedded Sensors and Actuators (ACESA) .,
program, and the Advanced Materials forApplication to Space Structures (AMASS) Zprogram, run by the AF Phillips Lab and (.""
Wright Lab respectively. The goal is to a
determine static limit load and fatigue jallowables, and to determine the effect oftemperature cycling on endurance and j, /performance of these members.
2. STATIC ACTUATION EVALUATION:The PZT d31 coefficient relates strain
40 -ii 0 0s
to field, and is a major determinant of &M,,NO..actuation and sensing capability in actively
Figure 2. Actuation Response of Navy Type 11
tDynamics Department PZT Encapsulated in Thermoset After 4000 ip-lMaterials Engineering Departn,'t Strain Tensile Load Conditioning.
D-7
more than the tangent actuation value when 0Each specimen was loaded in tension to a stressed beyond 2500 Ai-strains for Type II.
specified strain level. The specimens wereunloaded and the actuation curves werereplotted. The loading cycles were then 012,
repeated at higher strain levels untilsignificant degradation in actuation properties 0.1
were noted. The loss in actuation can be seen by 1 0.comparing the initial actuation curve seen in 0&
Figure I to the actuation curve of the samespecimen after a 4000 i-strain applied load, S O,_shown in Figure 2.
2.2 Actuation Definition: Most piezoelectric _.02
drivers employ PZTs in the d33 mode to apply 0 ,Iwo MWo 0W0 ,oo
axial point loads to structures. More recent WCR04nV1W OF DEECTO
smart structure applications employ the d31mode since thin wafers can apply loads in Figure 3. Computed Ejd31 as a Function of
parallel to thin composite structures. The d31 Applied Strain At Various Peak Voltage 1
value is determined by measuring the strain- Levels for Navy Type II in Thermoset.
voltage relation in the coupon and deriving thePZT coefficient using the relation: 2.4 Gr/Ep vs. Gr/Tp: Prior efforts were made
el = d3l (V/tp) (Elptp) / (Elptp + Elctc) to encapsulate the Navy type 11 PZTs into
where the subscripts p and c indicate PZT and graphite/PEEK composites. The high
composite properties, respectively. When the processing temperature of the PEEK 0composite stiffness is significantly greater thermoplastic (+7001 F) depoled the PZTs
than the PZT's, the product of PZT lateral resulting in a loss in actuation capability.
modulus (Elp) and piezoelectric coefficient is Additional thermoplastic encapsulation
more accurately determined. The material studies were performed using a lower processing
constant Ejd31 represents lateral stress temperature thermoplastic (Radel at 550* F).
generated per unit field, and is the b Figure 4 shows that the applied strain limit •measrted oer a it tiel auate fste bt a of the thermoplastic encapsulated Navy typemeasure of ability to actuate force on a 1 PZTs is significantly less (500 to 1000 ii-composite smart structure. sris hnta ftegaht/px
The d31 value presently quoted by most strains) than that of the graphite/epoxy
vendors is the initial or low voltage response of encapsulated wafers. Internal thermal
the wafer (note Reference). Actual d3 stresses, caused by the higher processinge ae not lier incse atemperature of the graphite/Radel composite,
responses are not linear, and increase as the are considered the cause of the reduced strainapplied voltage increases. The value of Eld31 limit.can be computed using peak strain and voltagelevels, or by using the initial tangent response.Figure 3 shows a plot of Eld31 calculated using 0.12
peak to peak and tangent actuation values.The bottom response curve (marked as 0) is the
actuation determined using the initial/tangent
voltage plots. The 50, 100, and 150 Volt peakto peak actuation plots are located above the .. ,tangent curve.
2.3 Definition of Strain to Failure: Static testswere initially performed on the ACESA 0 Sao l0o soo iooprogram and a tentative definition of failure %2:,104'M OF DEuECION
was determined to be 600 and 1,500 l-strains forTypes I and II, respectively. These definitionswere later redefined on the AMASS prgram. Figure 4. Computed Eld31 as a Function of 0This data, shown in Figure 3, illustrates that Applied Strain for Navy Type II inthe peak to peak actuation values decrease Thermoplastic.
D-8
3. DYNAMIC COUPON EVALUATION: Actuator/sensor feedforward is aFollowing definition of static failure measure of the actuator's ability to transfer
strains in the ACESA program, a set of four load to the composite and of the ability of thedynamic validation coupons were fabricated sensor to produce a charge proportional toand subjected to fatigue and thermal cycling, laminate deformation. The first three fatigueThe coupons were flat-sided tubes roughly 1/2 loading schedules were found to actuallyinches deep, 2 inches wide and 50 inches long. enhance actuator/sensor feedforward by 1-10%The layup had ten 00 plies and 5 sets of ±60. in all cases. Perhaps this is due to degradedplies of T300 graphite epoxy. Type I actuator stiffness of the laminate being pushed upon.and sensor wafers were 10 mils thick, and Type Schedule 1l-b, which was the only one11 actuator wafers were 15 mils thick with producing strains at 1,500 P-E, caused asensors at 7.5 mils. Wafers were poled so as to deterioration of at least 12% in feedforward.actuate and sense bending motions. Actuators Thermal cycling was found to be universallywere 12.5 inches long and were located near the damaging to feedforward, degradingbase of the cantilever beam. Sensors were 1.25 actuator/sensor transmissibility an additionalinches long. A nearly colocated sensor was 10-30% over that caused by mechanical fatiguelocated just below the actuator, and a colocated loads.sensor was in the middle of the actuator region. In summary, we feel that repeated
tensile mechanical loads on active graphite3.1 Fatigue and Thermal Cycle Testing: members is probably not a concern if strain is
Tensile fatigue loads were applied to kept below 1,500 p-E, and compressive loadsthe first two struts (1-a & l-a) at 60% of limit are of almost no concern. Repeated thermalas defined by ACESA testing. The lack of excursions over a wide range may cause anmeasurable damage led to the higher loading accumulation of damage, which warrantsschedule on stuts I-b and I1-b shown in Table 1. further investigation for space structures subjectSchedule b) is representative of cyclic tests to thousands of cycles over a lifetime.used for Space Shuttle fracture control.Thermal analysis indicated that passive 3.2 Thermal Testing: Capacitance andthermal control could maintain the dynamic transfer functions between actuatortemperature of a strut exposed in space within and sensor were measured in a thermalthe range ±100 ° C. Following fatigue loading, chamber at the completion of the thermalthe struts were subjected to 12 temperature cycle testing.cycles from -100. to .100" C. Capacitance of the embedded sensors
and actuators was found to correlate very wellTable]. Fati e Loadin Schedules with published vendor data. Figure 5 shows a
Type - Load Strain % of * of plot of normalized actuator capacitance forSchedule lb pa-e limit Type I and II struts, along with normalized
E-a 1 vendor dielectric constant data. It is apparentra1.0 that embedding the PZTs does not have a
2significant effect on thermal variations in2,000 480 80% dielectric performance. The Type I ceramic1,500 360 60% 800 to have a substantially lower
l1-a 4,200 900 60% 100 temperature sensitivity, especially in the
l-b 7,000 1,50 100% 10 range ±50 C.5,600 12 80% 50 Invariance of the transfer functions
4,200 90 1 6 over the operating temperature range is-- --- -important to maintain active damping
It was found that the fatigue loads performance, and stability margins of thedosed loop system. Figure 6 shows strut ['-a's
reduced f1 and 2 nd Mode frequencies an nearly colocated sensor transfer functionaverage of 1.2% and the thermal cycling measured with a charge amplifier at thereduced frequencies an additional 0.5%. This temperature extremes. The two curves areloss in stiffness could be due to a combination of shifted vertically slightly with respect toeZT and matrix cracking, fiber breakage and each other, indicative of changes inend fitting bond deterioration. Capacitance of feedforward gain. The second mode at 100. Cactuators and sensors was found to degrade an shows passive damping due to softening ataverage of i1% due to the fatigue loads, and an high temperature of the fiberglass supportadditional 1% due to thermal cycling, stand. The similarity in transfer functions
D-9
means that a suitably designed damping _
controller could operate without change over 1.3
the temperature range. Figure 7 shows a A ta-, /
comparison of feedforward ain from actuatorto sensor for Navy Type I and II validation 0 10(n]struts, using both voltage followers (VF) and A
charge amplifiers (CA). Type I appears to be -
slightly less temperature sensitive than TypeIt. The charge amplifiers give universally ..-- .. ..better performance than voltage followers. For .a charge amp the sensor/actuator gain term is zproportional to the material constants (El 0.9 -
d31)2, and for a voltage follower gain is S
proportional to (El d31 )2/D, where D is the o 0-100 0200
dielectric coefficient. The non-appearance of Temp IC
dielectric constant in gain is believed to resultin better temperature performance of thecharge amp. It was also found that the charge Figure 6. Variation of Actuator Capacitance foramp was not sensitive to shorts which Struts I-b and 1l-a Compared to Vendorsometime developed on the sensors, since the Dielectric Data (courtesy Vernitron).charge amp removes strain-induced chargefrom the sensor before the short can bleed it off.
The material constant Ejd31 is seen tobe a figure of merit for both sensing and o.o5 .
actuation.0.04,
.. ................ ....
ss~ I ' "
30
\-O 001 - _
c , N Figure 7. Variation of Feedforward froma0. Actuator to Colocated Sensor for Struts l-b and
_3,0 J .. .. .... ' w Il-a, Using Charge Amps and Voltage•i ,0 _ ; .. ....... . .Followers.
Developments in Piezoelectric Applications ofo -" ' \Ferroelectronics," Ferroelectronics Vol. 10, pp
£ "111"° u-11, Gordon and Brech, 1976.
-us0 ACKNOWLEDGEMENT: We would likei u' to thank the Air F:orce, and especially Lt.
Steve Griffin, the ACESA technical monitor,and Dr. Wayne Yuen, the AMASS technical
Figure 5. Nearly Colocated Sensor Transfer monitor, for their encouragement and helpfulFunctions for Strut lat Using Charge Amp. sgetos
-lO g
Preliminary SpecificationPiezoelectric CeramicsType I and Type II
* February, 1992
* D-i1
1 SCOPE
1.1 Scope
This specification establishes the requirements for thin (lessthan 20 mil thickness) lead zirconate titanate (PZT) piezoelectricwafers for use as actuators and sensors embedded in or bonded ontocomposite structures. Maximum encapsulation processing temperatureis 180 C. Service temperature range is +,- 125 C.
1.2 Classification
The requirements provide for use of piezoelectric ceramicco-ositions specified as follows:
Navy Type I- "Hard" piezoelectric lead zirconatetitanate ceramic composition
Navy Type II- "Soft" piezoelectric lead zirconatetitanate ceramic composition
2 APPLICABLE DOCUMENTS
The following documents form a part of this specification to theextent specified herein. Unless otherwise indicated, the issuein effect on the date of procurement placement shall apply.Later issues of these documents may be used at the option of thesupplier providing no degradation of the product ensues.Mandatory use of later documents shall be negotiated between thebuyer and the supplier.
SPECIFICATIONS
TRW Space and Defense Sector
STANDARDS
MilitaryMil Standard Specification of DOD Spec 1376A
Drawings
3 REQUIREMENTS
3.1 Electrode
Outer electrode composition is to be nickel deposited with anominal thickness of 0.0002". Silver can be utilized between the •
D-12 0
PZT substrate and the nickel coating.
Electrode placement is to be negotiated with supplier duringordering. Positive polarity is to be clearly identified oneach part.
3.2 Poling and Poling Date
The positive electrode is to be marked on each wafer with a rounddot or a plus "+". Poling date for each lot of material is to beindicated on the packing slip and on the packing container.
3.3 Dimensional Tolerances
3.3.1 Warpage/flatness
Waviness- Some manufacturing processes can result in awaviness or ruffled part. Unacceptable.
Flatness- A loaded flatness without cracking is acceptable.Apply weight to impart a 1-2 psi loading evenly over thesurface of the PZT coupon.
3.3.2 Thickness
* Thickness is to measure within +,- 10% of nominal. Thickness meas-urement is to be taken after electroding.
3.3.3 Length and Width
Length and width should measure within +,- .020".
3.4 Electrical Properties
3.4.1 Capacitance
Capacitance (C) is a function of thickness, surface area, and PZT* composition/manufacturing process. Nominal capacitance is computed
from the formula:T
Cnom- K33eoAJt
where KT 3 is the nominal dielectric constant, E0 is the permit-Stivity o free space, A is the nominal area, and t is the nominal
thickness. Measured capacitance of each wafer should be within5% of the mean for the lot. Mean capacitance of the wafers inthe lot is to be within 5% of the nominal capacitance.
E3.4.2 Lateral Stress Output Coefficient, Yjjd31
D-13
EThe quantity, Yj1d31 , is a measure of the lateral stress output perunit applied fi&ld. This quantity, defined here as lateral stressoutput coefficient, is an indicator of the ability to actuate uponstiff substrate materials. The stress output coefficient is alsoa measure of charge generated per unit area for applied lateralstrain. In this form it is an indicator of ability to sense lateralstrains of a substrate material when measured by a charge amplifier.
The nominal value of the lateral stress output coefficient is to begreater than 10 Newtons per volt-meter, regardless of composition.
3.4.3 Planar Coupling Factor, Kp
The planar coupling factor, K, is a measure of the ability to trans-form electrical energy into literal strain energy. It is stronglycorrelated with the lateral stress output coefficient defined in sectie3.4.2 and is easily measured. The nominal value of 5 at roomtemperature should be 0.550 or greater, regardless of composition.Variation of nominal p over the temperature range +,- 125 C is tofall within +,- 10% of the room temperature value.
4 Quality Assurance
4.1 Acceptance
The supplier shall certify that each PZT shipped for acceptanceconforms to the requirements of this specification and that allfabrication was accomplished in accordance with procedures, in-spections, and other controls found acceptable at the timeof qualification. Records of inspections, tests and othercontrols shall be available for examination by TRW for a periodof one year.
4.2 Responsibility for Acceptance Testing
The supplier shall be responsible for the performance of allacceptance tests and inspections, and in process inspectionsspecified herein.
4.3 Test Methods
Test method and sample size should be clearly identified.
5 PREPARATION FOR DELIVERY
5.1 Packaging
PZT ceramics are fragile and should be packaged in a mannerthat does not load the ceramic during shipment.
Each container is to be marked with the poling date of the
D-14
PZT ceramic along with its brand identification.
5.2 Marking
5.2.1 Individual
The positive pole of each PZT coupon is to be clearly marked withwater-proof ink.
5.2.2 Shipping Items
Each shipping container shall be permanently and legibly markedwith the followinq information which shall be visible withoutopening the container:
Part Number/SizeMaterial identificationPer this Specification NumberType designationQuantityElectrode MaterialPoling DatePurchase Order NumberManufacturer
m
0 D-15
ci,t:
EE
0 c0
in -a
0)0
E 0cn 0W
D-1
0E
I Um
II I I__ IfD
0 I D
cc
N 10
0 go C>~ 0
050E
ES 0
50 c c
2 z 0)0
0c 0U) 5 0
cm .2 E
0) 4) C~ O-
0 0i* 05 5
CUD 17
ii~ 0E wo
240-
U ) - o --
0) CU)
4) CL0
x ~ Cu
CC) 0 0.C
Q2 010
0 77
Cu Ch
uJ I
/o 2 -
QI 0) 0
+ +N
G)
G)) :E
D--
EE
00
*m o. 0En s:.0 V
*c Cc)t ~0Lw 0000)00 c. 00 Cc ~
ozm CL Em0 C
12 0 0 CD~
w 0 0
0 0 6
ClF
(M) c5500E~ ~~~i 0 05 0b5
cm c 0 0 0
.0E E *b.
0) -00 0.
0 0 c ;
~~U. .Z UC c C
E0
D-1
0I
O 0 00C0 000 0 0 c~o 0 0 I
JE~
* ) 0 Z 0)
a)a
* E EE0 w0 cm
C )Cwq 00
--
o~~c 0. jjL
4)
EA~CO UJ cc 'U
L..
II 0N
v
CC)
00C C
001&0
. - ft0 . IF o
p~
D-20
!E
cc 0.
i. U£ w -6 0
C 0 i )4
Tom~
Cu,
• Ei .= c
o ~ ~~ E= a-:"
4~~' 0 c 0 s-.
"~~L I:N ''
c1.
E In
v r.Ii, , ,I
05 o
N Q
00
CU A)
* ) 0.1)
N\ I
D-21
E
I II i I I I I I
.0
0 °
EE n
00
N - m
tn -0 0
Ni0 00 00o
00.
o , um00 Sm
0-
o., - .N 0
0-. a i EEl
II0
SEC., Ni>f oaco00 EN. o w 0 = cc * 0
II
D-22
UHiE
E CD
0 0
C.CLacu E
CL 0 00 cc
C) CD) 0 0
0 000C) 00 0
X 00 0L
E - CLE 0.-
1in0 0 r
C)- CO
Lu 0
0 c
a0 Cc r 00E 0 E'0
O~~0 =u. 0 :. E =.0 c0
So 0 0 >0( 0 0
0 0
S
u
D-23
N w
4EEA
D-2
z zZ
D-25
-oil
Ol-,ccwj2
(U)
c 0
D-2
00
* Uo
CA~O( Go 0 3'
~I..Gm3
S.. 0
oc.3C60 - "5 up
SD-2
C/)
0l)Iz 0
co 0
0 -
0 0
C/) 0 0
C 1a) a)
00I c' z
-4 w% -
60
E
0 E
D-2 8
x _xr1
r-4 C-) U) 0)C
00
N
c)
0)
C/)
Cec')
C) 0 C
L..L
C0)
SD-2
'o9
Cl) Z-1
Ocncc
D-30
00
C/) E o2a)
LN 0.
- 0czE a'a
LL 8
o00 04-A cUcO 1
* 0 mCJ
0 a
N fao 0.
0,
SCL
S 2 0
E E cm
r, -z
Ca 0) (IjCJCU
cc =0 0 0
0. 1
CD0
0 0 0,a?
0 3o
D-31
00
'A
-J 0
0 CC
inS
0- 0UC
D-32 3
0
- ____________________________________ * * I * I *
________ a -- ~-* C
* ___ I
0.9
o z CI I CI .9a
ftC I I I --
IH S* -
:~ 0 I 10~1I C
a.*~o pu~ - % ~*..I~ml
II~t I* -.- P----.-----.----.0
I II j
I t
~ : : a *C) - C) ft ft ft ft ft
(uzW) )usweogds~a (i'UWU) WID
* r I I
ft
II I I II I-1.-I-f- .ii 1 0
- I
~ Ii ~ : I* ii~i liii ~ -. ~ 1
I 00 K IDILLi -. -. -I I
I I~I .~ I
0m 0 W 0 'fl 0 m 0 m- - * - ft ft C 0 0 0 0* C W * .9 ft
9~ ft ft ft ft* ~ 0 (Muu)
D-33
in%
kNC
I II:
00en~ __ja n
-n -
(wiM) )uOsWsoWMt~ *Vi.OON
D-34
0
0
La La
0 La
U)
* U
<0 - z0 L
N w LU
D-35
.0
-0
0
000
0-36
*j 0
~~wi
14 -J
>~ U
nfl p
-d 00
* - *wC5i
D-3Z7 W
0
zz0
z z
IO~0 coooo
D-38
0z
0z
rA Z cn -21
co z
z
0 CA:;)
D-39
0.
D-40
ACTIVE MEMBER VIBRATION CONTROL FOR A 4 METER PRIMARYREFLECTOR SUPPORT STRUCTURE
J. W. Umiand" and G-S. Chen*Jet Propulsion Laboratory
California Institute of TechnologyPasadena. California 91109
successfully demonstrated an active damping concept forA hstraet both a beam-like truss structure1, and a more complex truss
structure2. This technique has been shown to effectivelyThe design and testing of a new low voltage damp structures that am truly free-fee 3, as well as those that
piezoelectric active member with integrated load cell and are rigidly 2 or softly attached to ground'. The activeisplacement sensor is described. This active member ismaximizing vibration
intended for micron level vibration and structural shape damping augmatin cist ofmmarimizin tonconrolof he recsio S~~meiedRefectr tst-ed.The damping by matching the active member impedance to thecontol of the Preision Segented Reflector tent-bed. The structue's impedance.
test-bed is an erectable 4 meter diameter backup supporttruss for a 2.4 meter focal length parabolic reflector. Active The present paper extends these studies to a 4-meterdamping of the test-bed is then demonstrated using d4 diameter primary support structure of a 2.4 meter focalnewly developed active members. The control technique length parabolic segmented reflector. The objective of thisused is referred to as bridge feedback. With this technique work is to demonstrate active damping on this test-bed. Thethe internal sensors are used in a local feedback loop to design and testing of an active member which utilizes a lowmatch the active member's input impedance to the strictre's voltage piezoelectric actuator, an integrated force sensor, andload impedance, which then maximizes vibrational energy a simplified elastic deflection sensing scheme is described.dissipation. The active damping effectiveness is then Experiments are then performed to evaluate both the newevaluated from closed Icop freqluency responses, active member, and the effectiveness of the, control strategy.
The active damping performance is evaluated by comparingthe closed loop response to the open loop response.
One major element of future spacebrne asromicl Damping performance is demonstrated using: (1) an
installations will be a structural system with strict individual active member and, (2) two active members
dimensional accuracy requirements. In the case of a operating simultaneously.diffraction limited telescope, the wavefront error will be on Test-bed Structurethe order of X/10. where ) represents the wavelength ofinterest. Based on dhs wavefrom error, the surface eo will The dynamic test-bed, shown in figure 1. is made ofBe )/20. Given a surface error budget of ).120, flight-like hardware consisting of 150 graphite/epoxy trussapproximately ninety percent of the surface amo is allocated members and 300 erectable aluminum joints. The totalto the primary reflector. Therefore, the primary reflector's weight of the structure is 86.4 kg. This structure wasrms surface accuracy must be kept to within a fiction of the designed and frated at NASA's LaRC'. The geometry ofobterving wavelength (e.g., 1/20 - 1/25) during the this highly redundant construction is driven by the locationobserving period. Therefore, for an infrared astronomy of the nodes on the upper surface, which are determinedapplication, the structural accuracy requirements are on the b on th desied refle= pawl locations. The nodes anorder of a micron. Vibration arising from transient dynamic the upper surface are then connected by members of thedisturbances generated f o on-a- d equipment or e required length. The lower surface nodes are located byoperation is one potential source of reduced structural reuedlnt.Teowrsraeoesreocedbforming a tripod of equal length members whose base isaccuracy. and hence telesimpe pefarmanm determined by thee adjacem nodes on the upper surface. The
At the Jet Propulsion Laborary (JPL), the Precision lower surface is then formed by connecting the nodes withSegmented Reflector (PSR) program has investigated an members of suitable length. The stiffness of an individualactive structure approach to mechanical vibration core member is 19.SN/Iln. The stiffness of the upper andsuppression. The active structure uniqueness involves the lower surface members will vary according to memberuse of stauctural components in which actuat, sensors and length. For the purposes of this investigation the structurefeedback control are integrated into an active member. is rigidly fixed at the three central noe of the lower surface.Using this approach, previous investigations at JPL have A 1130 kg steel block is used as ground during tsting.
"AlAA Member,DynaicsRch Labonory,Applied Mecamics Technologies Section.Copyright 0 1992 by the Americacn Inate of Aeronautics and Asnrnautics, Inc. All rights reserved.
D-41
yow
Figure 2: Erectable active damping member.
Two active members, whose major componentsconsist of an integrated piezoelectric actuator and loadsensor, a displacement sensor, and erectable joints, werefabricated, see figure 2. A detailed assembly dawing of themember is shown in figure 3. While these new active
Figure 1: PSR TB - members are similar to those used in previousinvestigptions 5. 6, there are sifcant differences, namely.(1) a low voltage piezoelectic material is used, (2) the load
The structure's dynamic charactexistic am detailed in sesor is integaed into the same peckag as the mOtor, andtable 1. Now that the first three modes can be described as a (3) the displacement sensing scheme is simplified.relative shearing of the upper surface with respect to thelower surface. The fourth mode is described as a breathing The design challenge undertaken here was: to producemode. As the aspect ratio, relating the support structure's an active member whose performance was equal to or betterdiameter to thickness, is increased, as would be the case for tha previous designs, was simpler, and placed less stringent 0the Large Deployable Reflector (LDR), these breathing requiremems on exrnal components. In the previous activemodes would become more predominate. Therefore, it is member designs, a high voltage piezoelectric material wasdesirable at the current test-bed phase of the PSR program to used. This actuator material produces a 63 micr..,demonstrate an ability to control this type of mode. displacement when 1000 V is applied. The ±31 micron
operational dynamic range of these acmamrs was achieved byTable 1: Test-bed dvnanic characteristics. applying ±500V with a 500V bias voltage. Here, a low
Fjj. (Hz) Damping voltage piezoelectric material was chosen such that theMode LaRC Fl. (HZ) Raio % operational range of the actuator was reduced to±75V with aNo. Data4 JPL Data JPL Data Description 75V bias voltage.
In both of the previous designs the piezoelectric stack1 4.5 32.a 0.9 x Rocking was pre-loaded by a pre-load spring, such that the ceramicShag meral never experienced a tensile Smss. The moe recent t
active member design 6, used a pe-load spring and parallel
35.6 33.4 0.9 y Rocking motion flexures to eliminate the 'stiction' observed in theand active member5 pre-loaded with a Belleville washer.
Shearing Futermore. crossblade flexures were incorporated into the
3 51.9 42.3 0.7 Twisting PARALLEL MOTION
4 57.3 56.0 1.8 Twisting CROSS SLADE SENSOR OUTPUTad FFLEXURE TARGET-\ STEM
5 78.1 75.8 3.3 Istz Breathing
6 96.6 93.4 1.1 z Saddle PZT ACTUATOR/ PRELOAD NUT
LOAD SENSOR SPRING_; T
7 97.3 97.6 1.5 z Saddle DISPLACEMENTSENSOR
ID@ c~r~iliinn, Figure 3: Assembly drawing of active member.
D-42 0
Load Sensorwafer Table 2: Piezoelectic stack parameters.
M insulating Wafer PZT water diameter (mfg data) 16mm50 mm
F 5 M PZT wafer thickness (mfg data) 0.8mm
Number of PZT wafers in acwuator 66(mfg dam)
SINumber of PZT wafers in load sensor I
Electrical Insulation g33 (mfg data) 0.023 VmN
-.- 70.mm93 (m gd t)00
3V /
J g ACuator Sainless Steel i33 (mfg dam) 530 pC/Nwaers-. Eno Caos e dielectric constant (mfg dam) 2600 Flm
Figure 4: PZT stack drawing. Stiffness, short circuit 64 Nln
active member design 6, to isolate the actuator stack from Stiffness, open circuit 99 N/pmIrdmng moments. The flexure and preload spring design ofthe JPL active member 6 is also incorporated into this Caacitance, actuator 2 F
Cavacitance. load cell 3.5 nFThe design of the piezoelectric stack is shown in
detail in figure 4. The actuator is built up by stacking The displacement transducer is the same as usedwafers 16mm in diameter and 0.8mm thick. The wafers am previously- a Kaman-7200 eddy current proximity sensor.bonded together with an elecrode placed between each wafer. In the previous designs5.6 , the motion sensor was referencedThe electrodes are wired in parallel, such that the voltage to the fixed end of the active member by a reference rod.applied across an individual wafer produces an electric field The actuator stack was annular, so that the reference rodin the appropria direction for the wafer. The underlying could be passed through. A different displacementmechanism of the piezoelectric actuator relies on the measuring scheme was used here, where the motion sensorelectromechanical coupling of the piezoelectric material. In was mounted directly to the active member casing. Theother words, when an electric field is applied to the casing of the member is not in the load path, i. e. it doespiezoelectric maerial a mechanical strain related to the field not deflect elastically, thus leaving the motion sensor fixed4s created. This relationship tends to be nonlinear, and relative to the dead end (left end in figure 3) of the member.11Stftic. The motion sensor target is mounted on a yoke, which is an
Included in the same package as the motor is a single integral pomon of the acte member output stem.
wafer of piezoelectric material that is used as a load sensor. The cross-blade flexure located at the live end (rightThe same piezoelectric material that is used for the actuator end of figure 3) of the active member is an integral portonportion of the stack is used for the load cell. This wafer is of the output stem. The live end of the assembly iselectrically insulated from the actuator wafers. The supported laterally by a parallel motion flexure whichoperation of the load sensor relies on the convene of the consists of a pair of wide blade flexures. The cross-bladepiezoelectric effect. L e. when a pressure is applied to the flexure, piezoceramic stack, and output stem are placed intopiezoelectric material an electric charge is produced across the active member casing and held in place with an end cap.the material. The end cap is mated to the active member casing by a lens-
type scew thread. The desired compressive preload on theThe panete of the piezoelectc stack we listed in actuator assembly is then applied by a preload spring andtable 2. Not, that as expected the stack open circuited nut.stiffness is 55% greater than the short circuited stiffness. Itis generally expected that a PZT is roughly twice as stiff A 140 lb compressive preload is applied to the activeopen circuited as short circuited6 . Finally, stainless steel member by using an Instron machine. The preload nut iscaps are bonded to the end of the suck. and RgV insulation screwed into the active member end cap. such that it doesis used on the oumr smrace, not contact the preload spring. The assembly is then placed
into the Instron machine. End ftinus and ball bearings areused at both ends of the active member such that bending
D-43
0moments are not applied. The active member is thencompressively loaded until the Instron machine load cell 4.0-indicates 140 lb has been applied. The preload nut is slowly .s,, .82 vI ....................Lightened until the load cel reading just begins to drop. . .........................'. ............................... . .. ..........This indicates that the preload nut has just contacted the -p~2. .od , ~ g n h oa n te z s .= , ....................... ....................... ..... ...................preload spring. and that the load on the Instron machine is LL........being relieved. The external load is removed, and the active a 2.0- ...................... .............. ................member assembly is completed. L.5- ........... ............................
-L Active Memher Characteriatin ad 1.0. ....... ... ...........
........0.5 .......
1 2 3 4 5
The effectiveness of the active member, and the low PO La 5I)
voltage material can be evaluated based on a series ofexperiments. First, the active member sensors are Figure 5: Dead weight load cell calibration curve.,alibrated. Secondly, the effective piezoelectric field relationof the active member is evaluated. The active member applying a known dead weight load. Sufficient time wasstiffness is measured. -- allowed to pass, such that the signal produced from the
system was zero. The system was trggered. the dead weightFor all tests where the actuator was drivei, a Burleigh was quickly removed and the resulting signal was recorded.
PZ- 150 power amplifier was used. Unless it is otherwise The load cell sees the removal of the dead weight as anindicated all the data was acquired using a Tektronix 2630 effective tensile load. and the peak value is correlated to the4-channei Fourir anay-e,. load removed. Unfortunately, for this system the AC
coupling is relatively fast and this technique producesL-.- igfn ,2 eJ a n Sesn-gr Ca1ibration significant scatter in the results, see figure 5.
The internal eddy current displacement sensor was The load cell was also calibrated with an impactcalibrated once the active member was completely hammer. The stack was placed vertically on a rigid base,assembled. A Schaeviz LB-375 Linear Variable Differential and the free end of the stack was tapped with a Dytran 4122Transformer (LVDI) was used as the reference measurement impact hammer. The signals from both the impact hammer.The active member was fixtred such that it was horizontal, and the charge amplifier were then acquired with a Fourierand its dead end was fixed. At the live end of the member analyzer and an averaged transfer function was calculated. seethe LVDT was brought into contact with the member. The figure 6. The sensitivity of the load cell was then calculatedLVDT was adjusted so that its axis was parallel to that of based on the magnitude of the flat portion of the frequencythe active member. The internal displacement sensor was response. from approximately 10 to 100 Hz. The effects ofcalibrated dynamically by driving the actuator with a 1Hz the electronic AC coupling are evident at low frequency,sinusoidal voltage. The signals from both the internal and where it is observed that there is some amplification, andexternal displacement sensors were then recorded phase lead. 0sinulunously. The sensitivity of the internal displacementsensors were then determined to be 2S.3utm/V. The eddy A." Hysteresis:urrent displacement transducer displayed zero hysteresismith respect to the LVDT. The displacements produced by a I Hz sinusoidal
voltage input to the actuator are shown in figure 7. AsLtL LAad QALL £,2k1a1r na
The internal load cell was calibrated before the active 3S .......................... 200nember was assembled. An lbstum 2606 charge amplifiervas used as the sipal conditioning for the load cell This . .... ,_
re amplifir is AC coupled via a second order high pam I __c,
il er tha has a cu-off fr uency of 0.5 Hz. ad a damping 20 0- 100 HE vIM .wo of0.7. - _______ . -
The firt calibration technique used is the standard i ")ad cell calibration method utilized by the JPL s5. .... .....Ist'umentation sction. The data was acquired using a 3 a --/r---' -------- . 0igiazing voltmeter and ransferred to a computer system for 1 8 * a * 4 1 a 3 * 00
rimt out. The system was mggeed manually by removing _
jumper. thus enabling data acquisition. The load cell wasdibrated by aligning the integrated sack vertically, and Figure 4: Dynamic load cell calbrtin
D-44 0
... , . .0 ". ..................... ...... . .-" .."..... .. ..... ...... . ............ .
1 0.. 260.... . .---------: . ........... ... . . . . . .. ............ .. .......- .....................
: .54 24 . -........ - .. .. ...... .................. . ........ "
2 - Q .- - .... . . .--- .... .......... HZ .. .- --. -............. ..................
S2 0 5- ." 0 - - ......
i I I I i i
0 40 so 120 0 10 20 30 t0 50
ADOa V i AA Peak-pea Oisre pmlwt fun)
Figure 7: Active member hysteresis. Figure 8: Active member gain.
expected the actuator displays some hysteresis. The plot in increasing gain, is that for control design an increase in pinfigure 6 shows four different hysteresis loops due to four reduces the stability margin.different input voltages. A measure of the hysteresis presentin each loop is defined as the aspect ratio of the loop. The An aging effect was also observed. The gainaspect ratio of the loop is the maximum loop displacement measurements were measur three separate res, for a 1 Hzwidth divided by the peak to peak displacement The percent sinusoidal input. A noticeable decrease in gain over timehysteresis present in the four loops is 83%. 10.2%. 11.3%, was observed, see figure 9. This can be detrimental toand 10.9%. These hysteresis values are stated in order of contol system performance over time. since the system willincreasing peak to peak displacement. The hysteresis shown tend to detune itself.here is a reflection of the lossy dielectric coefficient of thepZT. . S± Li--
A nonlinearity, such as hysteresis, in an actuator can A series of step response tests were performed on thetend to be detrmental to the actuators performance. For active member. The step response shown in figure 10 is theexample, if the actuator is to be used as a displacement displacement response of the active member given a 115driver a control system that will eliminate the nonlinearity Volt step input. This plot shows the quick responsemay be required for accurate and repeatable operation. In capability of the PZT material along with its long termoter words the hysteresis can produce residual displacements creep. There is some initial oscillation due the imperfectat DC. step produced by the function generator. The response
reaches 95% of its final value due to the initial motion ofL La the actuator. Approximately 25 seconds are required for the
-. response to reach its final value. Qualitatively, thisThe fundamental field strain relation of the PZT is behavior was observed for the 33 micron response of figure
also indicated in figure 7. That is, for a voltage applied to 10, and for step responses down to 2 microns.the piezoelectric material a propormonal strain is induced.Foe larger peak to peak displacements the slope of thehysteresis loop increases. This implies that the mateal'spiezoelectric coefficient increases with increasingdisplacement. An actuator pain can be defined as the raio of ,,the peak-to-peak displacement to the peak-to-peak applied H19192voltage. The actaor gin has been calculated for the °,0hysteresis loops of figure 7, a well as for several otherdriving frequencies. These reatlts are indicated in figure 8. 2This data shows that fhe PZT pin incrases with increasingampfitude, and demue with icrsng driving frequency. 20- 11/_22,221
The behavior of the PZT gain increasing with 240- 0 ....increasing displacement is somewhat analogous to a frictioncoefficient versus sliding velocity curve with a negative 220 .......
slope. In the case of the friction versus velocity cuve, the 2 , a , a ,faster the motion the less resstance there is to that motion. ,0 20 2o 40
These two cases ae analogous in the sense that they both - o'g FIRM)
have a destabilizing influence. The difficulty with an Figure 9: Active member pin aging.
D-45
0
I -numerically differentiated to give an impedance. Thisimpedance will be highly frequency dependent due to thedynamics of the structure. Secondly the open loop active
"0.. . member input impedance is measured. This is done by... ........ measuring the same transfer function as above, but with the
...... . structure being driven by an external disturbance. Typically.., 90 ,C 20 the active member open loop input impedance will be notf0 40 frequency dependent and is usualy an order of magnide less
............................. _•__,__ ; - than the average value of the load impedance. Although, theactive member impedance will have a positive slope since an
_ _...... open loop active member is essentially a passive spring.C " ... This implies that the active member is stiffer than the 0020 40 so 8
-rif i suuctre. The feedback paths are then closed and the activemember impedance is again measured. The feedback gains
Figure 10: Active step response. are adjusted until the active member impedance is-. approximately equal to the average value of the load
impedance. The active member is electunically sofwned by4 Active Member Stiffes this procedure.
The active member stiffness was measurmd using the Active Member ZpaegmetutInsm machine and the same fixturing that was used duringassembly. This test is static in nature, therefore the active An objective of this study is to demonstrate themembers AC coupled internal load cell is not used. Rather, ability to control the test-bed's first four modes. Since onlythe Instron machine load cell is used to measure the applied two active member were available, it was decided to performforce. The internal displacement sensor is used to measure separate tests to control individually targeted modes. Given 0the resulting deflection. The average open circuit active the small number of actuators and the experimental scenario,member stiffness is 44.1 N/un. and the average short circuit the active member placement is relatively straightforward.stiffness is 39.8 Nun. The stiffness of the active member Based on engineering judgement it was decided to replace theis different from the stiffness of the PZT stack because the two structural members that saw the highest percentagestiffness of the crossblade flexures is in series with the strain for the mode of interest with the active members.stack. The preload spring is then in parallel with the Since this is a smcture with clamped boundary conditions, *stiffness of the PZT stack and flexures. it was intuitively expected that the struts with high strain
energy would be near the mounting points. Using thisLAetive Damniny Aurment tion rationale the active member locations were chosen, and are
shown in figure 12. Note, for each of the chosen locationsThe damping augmentation concept used, is described one end of the active member is attached to ground.
in this section. First the feedback control strategy and gain Secondly, the active member placement for the third mode is-adjustment is summarized, then the active member on the lower surface rather than at one of the core memberplacement problem is considered. locations, as is the case for modes one, two, and four.
14Feedback Control LAX
The feedback control straegy used in this e.xperiment PERFORMANCE
is referred to as bridge feedback l , where the active member's 0 I ---- ---ES
force and relative velocity signals ae fed back locally. This (STRUCTURE)control strategy is summarized in the block diagram shownin figure I I. The advantage of using bridge feedback is thatthe desired loop gain funcion wid the active member's input ACSPt-ACEOENS R Limpedance function are implemented without adverseinteraction and conflicting requirements. This control CIEM ,strategy is then implemented as an analog circuit.
The gain tuning strategy used here is as follows. DEFORMATION FORCEFirst the lId impedance function is measured. This is doneby driving the structum with the active member acuatm, andmeasuring the transfer function between the displacementsensor and the load cell. This transfer function is then COMPENSTR
Figure 11: Bridge feedback control block diagram.
D-46
Z 1 ; Md. 3Mode4Mmde I Mode 2 Mode B103mn
X Rockin Y Rocking Torsion E 103XE~om97 E61 E 105
E 102 E 100 X
- Dma. Aciti Momb Lom a
Figure 12: Active member placement
outer comers, as shown in figure 13. Dead weights wereL L ia mentaL Rlts placed at the remaining two upper surface outside corners to
simulate the mass of an accelerometer set. A mini-shakerTeest-bed snucture was instrumented with twelve was used as an extemrnal disturbanc source.
Stmdsrand QA- 1400 servo accelerometers, grouped into fourtriaxial sets. The accelerometers were AC coupled such that Li1 zrthe constant acceleration due to gravity was removed fromthe vertically mounted accelerometers. The accelerometers All test data was taken in the form of transferwere placed at four of the test-bed suucure's upper surface functions generated by a broadband random input signal.
The active members were placed in the structure and theI 36 1303 gains tuned according the technique described above.
Transfer functions of the acceleeometer response to the inputforce were taken with the active damping disabled. i. e. openloop. and enabled. L e. closed loop.
1308 1301
Inially, open loop response tests were performed oneach feedback channel Transfer functions were measuredbetween the compensator inputs and the active memberinternal sensors. A negative stability margin was observedat approximately 2kHz on the force feedback loop givenonly modest gain levels. Normally, force feedback of anactive member has 180 degrees of phase margin 2 . Thenegative stability margin was caused by the combination ofthree factors. Firt,. the compensator roils off at -4dB/octave
Y with a phase change of approximately -30 degrees.Secondly. the power amplifier coupled with the capacitanceof the actuator act as a second order system whose naturalfrequenmy is 900Hz and damping ratio is 0.7. The amplifier
1313 and actuator system produce a 180 degree phase change atS210 2 kHz. Finally. the active member internal force transducer
* - 5ucoort Point observes an axial resonance at 2 kHz that has a highdynamic amplifrication. It is because of this axial resonance
0 - Triax Accelercmete- Location thatanegasive gain marginisobserved.C)- Dummy Mass Location
Figure 13: Experimental setup.
D-47
. ....... ......
..........
=T ............... . .... .....
I.~~.... .....L ... I
S100 .D 9!................
... ...... ......... .... . . . . . - .. . .
..0 ... . ...0 . ........ ....
... ..o..nt..d a...... ...e....p.of....e....tive.member ....e .oa ..idivn 1 0 06
.......... ....... . ...... ...................... .. ........ . . ...... ....... .
Figure .. 14 A..........v .... m.mbe...ds..u..re ...e..c. . Fi u e1 6: Op; an .lse l .o.... ~ u externa2IOXY. In ..gure 1.. .o... oo
onde oersrf30 o h stucur t0 prodda gto ue to the moa naua feunis anamigrai0
thge 1sina Aciemod a d expectrte lopanimeda. Thge ampingnrand clroted ls osona mod-e is improved
dnotdin thoem igras bysis g rqun eenldet aome0.7%metor res0%.s as note thatX lbdfue n fotrnthicepresented g the yam of the tucemr. The oploop molldeiin fasrcende slightyI fromr 4215 he ope 41.opz
~~~~~~~~~~~~~~active member imedne.deoed asZ ~loi iialacmain o the structures ope and clsedinAne load mpne. Aditoaly the active member 16.dnc Hmod e response ofplte atm toroalod is reduced mo
macigt h odimpedance is raulyicesing in frue. ThAn b a3ac. ofil 6. amlteo dampongeraofti mode rmis
inoctes tha the opguel astv meber is frqeyectieydeicree from 0.7% to 2%. wnoehile the frequency hassprening thatitfer thnamic ofe srucue. The on cloed mdrped haeueshl.fom 42.1 Hz to 40.1 Hz.loactive member impedance denoted a cl3oe loopi Siial.acmrsooftetutrsopnndlsdinougure an isde apoxmantue equa ta the average value of The -s atve member weratie thnen usown repac cogrethe load impedance. Theclodlo, te activmembers ofHe the sne he acstatorl o psrodeimedvate proceaduall otinedaboveg wthe reure Tkes byafcoof6Th damping totefiraw no.tie fortis mod is
onlyiaafew iherationsountiloth active member impefedaney a incresd byro n a.7 o cal.wle the frequnc Th Tspintai ste whan give at siac defeeo byne clogseg arpe 10o lb. dea weightHzloo fromv nodee 210.ace Thete as was thend exltdoysudelThfiue JAiasfrofunctons shown in fige Sa showlue cutin the scting th eldth weight.use thereipartingra
ft ~ ~ ~ ~ ~ ~ ~ _________ lodipdne0I lu opfebc an mmwo h t i ctt ct rvd
Indvatepoeueodndaoe h rcd M dmigt h is w o= ITiw~m dmaidawsonly~~ ~ ~ ~ a eItrtosutltead ebripdnei aqie ypromn ocle tag m7iT
optimanyo UUNXL w~as given a-ttcdfeto yhngn 0l edwih
.0.20- OM f
30 .... 5.00 60I IA I. 0. .01 .
Figure 13: Open and closed loop structure frequency Figure 17: Time domain response, 2 active membersresponse functions using I active member. placed in the TB cam~, control off.
D-480
0.30 LAcknowliedgcetnts............ ................................ .T. ...
0.2.Two A M"" .o . The research described in this paper was performed atS. .......... ................................................................................ the Jet Propulsion Laboratory. California Institute of..... .................. ............................ Technology, under a contract with the National Aeronautics
....... .. .... .................... arid Space Administration. code RM. Thbe authors gratefullyS.oo 10 ... acknowledge the assistance of the following: Chris Miller.
......................................... ..........}.....-.".... John O'Brien. Andy Kissil, and the JPL Dynamics• Laboratory.
.0.20- 9. .L~z~~~0.0 0.1 0 Rfr 20
Time ,,ecl [1] G-S. Chen and B. J. Lurie. "Bridge Feedback forFigure IS: Time domain response. 2 active members Active Damping Augmentation." AIAA Paper 90-1243.
.laced in the TB core, control on. 31st AIAAIASMEIASCEIAHSIASC Structures. StructuralDynamics. and Materials Conference. Long Beach, CA,
sitep change in applied load. The open and closed loop April 1990. also to appear AIAA 1. of Guidance. Dynamics.acceleration response at node 1313 in the z direction is and Control.
0 shown in figures 17 and 18. respectively. Note that withthe control off, the response clearly shows a low frequency (2] J. L. Fanson, et al, "System Identification andmode of approximately 32 Hz. It is also evident, because of Control of the JPL Active Strucre," AIAA Paper 91-1231.the beating phenomenon, that there is a second mode that is 32nd AIAAIASMEIASCEIAHSIASC Structures. Structuralof roughly the same frequency. These two modes are the Dynamics. and Materiais Conference. Baltimore. MD. AprilTB's x and y rocking modes. The TB's settling time with 1991.the control turned off is greater than two seconds. When thecontrol is turned on the settling time is less than a half a (3) C. R. Lawrence, B. J. Lurie. G-S. Chen. and A. D.second and the low requency modes are no longer eviden. Swanson. Active Member Vibration Control Experiment in
a KC-135 Reduced Gravity Environment" First USIJapanR Umm a2r Conference on Adaptive Structures. Maui, HA. November
1990.In this paper the design and testing of a new low
voltage active member was deschbed. The maximum s [4] H. G. Bush. et al. "Design and Fabrication of anof this active member was measured to be 45 microns, Erectable Truss for Precision Segmented Reflectorwhich is more than adequate for vibration damping Application.' AIAA Paper 90-0999. 3 1 s tapplications in precision structures. The internal AIAAIASMEIASCEIAHSIASC Structures. Structuraldisplacement sensing functioned as expected. The internal Dynamics. and Materials Conference. Long Beach. CA.-force sensor displayed an active member axial resonance that April 1990.had a very high dynamic amplification. This resonance ledto a closed loop stability problem. This problem was (51 Kaman Aerospace Corporation. Tonstant Lengthresolved by using an external load cell. The nature of the Strut." United States Patent No. 4.742.261. May 3. 1988.actwe member axial resonance, and compenSating for it is asubject for future work. [61 E. H. Anderson. D. M. Moore, 1. L Fanson and
M. A. Ealey, "Development of an Active Member Using
Active modal damping was succesfuily demotrated Piezoelectric and Elecrostrictive Actuation for Control ofon a realistic dynamic test-bed. The low voltage active Precision Structures," AIAA Paper 90-1085. 31stmember, using only collocated force and displacement AIAAIASMEIASCEIAHS1ASC Structures. Structuralmeasurements, was dynamically softened by electronic Dynamics. and Materials Conference. Long Beach. CA.feedback such that its impedance matched the load April 1990.impedance. The benefit of impedance matching is that thevibrational energy dissipation is maximized. The advamgeof bridge feedback control, given the large stability marginsnormally associated with force feedlack, is that it is robustwith respect to the uncertainty of the structure's dynamicchmnmcs.
D-4 9
0
NOm
a:0
CcU> C0
Cc CD
ra
00L
0-
0
D-5
C.)OIE
0m N N
CL E
CL cm
IM 0 0
0 00 0 MN N
ui ra 1. E0. CL 0 0-s - 0 0
0 ZCl) ~a CLI L
ui 0
aE 00 ~0
CL)U) U0 0-
E 0CU
oL o E, oLo Cu
CuC
C/) a~a.1
D-51
C
0 00 CN
00
50 EiE
z Uwi 00
&- OC
) *Coo 0
ZW 0
LmL 0 0,*
C')- 1wlec~m 0
wcnv-v.-w V- IV V
m c00*
0iD-5a
zLLII
L))w% u -cm Ew m/ -u c
ccW oA (ni-I0 _.C)
0 w -jS0 ) 0. - P
z0 z0 _-W > r
C/) p wq4) D
00 cc
U) z I I00 (
CC.)
-IU 000*
0 U. 12 0~ ,0: E
Z cc6 - 1V) V( 0 to
. -'r
0q CDC0 ED1=3
00
Q.00
cc
0.40
CD
EE• - C% N
CC, ,) Ln
N l" 0 Z: A
0,v1to - U) C 00o (n U o
E 0
a 0 •. -0l 0¢ 0. L.AM 0.go L 0. 0 0E E)
0 2 U) U
0 0 omi ONO
- •
D-540
I.PI
CA55
E*je
2 01
CO CL
E
D-55
cVEE
0 0
U)
I- U
Lflm UU)
0
U)) lz
xx
I 0 ( 0
*4 0 0h 0
U)(
c U) 0
m-56
zz00
o czC,)-
wU zLU
I-r0zz
U) .)
WW- C) C-,
C/) c-
'0LU C,)00 0
~LL <( 0j 0d<= Cl) C
ZZU >mZ 4c LU 0 12
0 U~~~C) Ca<>-i c
LUU)ZZU >-z WU) 0
0z no a. Lu zi c. cL
Lu P. c o mUU
M LU (nLD-5 4
00
C0
w 0
0 #0-
a)Lam
0 201wE 0
SEM 0.
00 a) a)
a)m 0)0o ho.ur4000000
a& 0- o
0"WIN::~
- ~a~aD-58
-kE
LU00
r zz1) I
U, CD
0 cc
D-59
00
ZZ
LU
Eu w0>
4J 4
..1o ...0
0& 04C1 0 us
U-60
W 0
IL I I9
IinC))
LL 0 >~
C.) '
.
I-t00
00
z
LCL
D-61
C 9
I- Kii~ij)14'(El) =
* L L -CL
(5-
w/
/ a)
I-A0'
L U4.
0~0.O.0-
CL
4-
cc
00
D-62
LU Ln22 1 EC
ul go
Z %goo0 41 IL c
C013 LL- I
am z SO C
uj U)la- U
=+- z
> U..100ILp 0IZ
(M
OD-6
pcl
OWN C
z IC
El '.j* i
L~g
l z, - NClI0I< .1zoEnIE . 2 W
U 00
zCo z
IL~ccw I.1
D-64
0
00
o ==CD c
E~~9 Cl)
U) L.
0) oo ra) =)
E -E~ ca~ w~)M) CO E L
0) Eo ci w
CL 0 ) >~~~a(
CDO 0>CDE 25 8 E r- ) c .2~.0
0.0 (a E > oZ5 C C.-
-J)
cc 0 A CL s-D=-53
A ~ . '
A
.9
M* I**
11w -JI
h
D-67-
.........
.. . .. . ..-
rr) .... .. . ......
________________. ....
.....
:7 7-L::: 7-7z:. :
-i 1V~-- ---------Z" I. 77 T
7: WA 74:
7 :17
I
I
LL~i t.It. i I IF II-
- T I I I I iI T I L
I I*i
14-419
a x
I L I
to-
Um
C. I 1 II •
D-70
.. .... ...
.1-
.LL LC=~
-JL
C) LU.C/) C=*.
LL. GO
I- LUJ LQ- LL c
I L LU--l LU -
_ .. I-LUJ LU am..Z n -
_ ~ U LUJZ3
LA C- )
C=LLU LUJ L..J 6I a- g.MCp .- I I-LU
LUI LUJ r- D
LUJ CDC LU LU JC> L- LU I I- LU I
a=2: eLUL h coC:
C)- I- CD. $- --
C:) -JJ I-
w) C) . -l c.D C) CL= leC- L- = - I
rl%4L< LLJ LU LU PUCe I-- c..) I-- = U.. 1
C/) Ln LUJ .J 0 -C/) I-- =, LL- 3cI-IL ac) C- 4<z LJ LU
C)
D-711
z0
z +# Dh(UaM q-
0In 4)
Iwoin
wmw
mmml 0 U)
0m00
0
D-729
0-
I-l
CL
D-73
oo
aC
I.. 4.0 F.
0 z It
~ N > Woo
CL
Cb c 0am 0 cm a
c D-7c
.........
-n () W)m
.O c inU 00S
C14a x . a C4 > 1-- 00 11 1 U
U .woU) E Un
co 0 co 0-1
> 0 S~~ ~
Cl)C )O.,..2 CO0
( ) " o .. . e v . 9ER U , 0
EE
0b 0 0 w0 3:C o
> U,
N. 00LL ,-.
-u 0
%A JAl. a +E . LJ
N~
D-75
(STUNW) SSNADINI
NY 0D co %0R0D
* (S.LVM) H3MOd
01, %
Sc
tnU
Ci) 0 0'm2t
-J)
z
0
CC
-IL
(NI) 1H9)13H
D-7 6
- - --
E-.
C .
OrJ C "-
D-77
E0
ON
00
I-
cc A oo
04• c 0 I O E
-- F ) 06 .,
0 q oP "
0 0
"C-) -- .-In 0. U) 0 E.
>0. g 400 . 0 Z3: ~~4~- 0.0000C .C
EO =) *.0 0. o .0c0 4w9
Co<
'"0' m 0 0 -% j
> >
D-78
Workshop on Advanced Piezoelectric Actuator materials forSpace Applications, Arlington, VA ; Februarary 25, 1992
APPLICATION OF PIEZOELECTRIC MATERIALS
TO MECHANICAL RESPONSE CONTROL
Vijay K. Varadan and Vasundara V. Varadan
Alumni Distinguished Professors of EngineeringCenter for the Engineering of Electronic and
Acoustic Materials
Pennsylvania State University, University Park, Pa
D-79
Ills'
D/80ii Ii , •
!111•
'I
D-80q
SOL-OE.PREPARATION
ULTRA-PURE
4 .. GRAIN SUZ
PREPARADUCN
D-8
Starting powders
Mix powdersand binders
Tape Electrode Lamination
Cast tape m
Cut _
Screen printelectrodes •
Burn outbinder Temierature profile 1 hour
Sinter Binder bum out .600
43 dayTerminate
4 'W 40 4 Min.Attach leads
Encapsulate
D-82 0
> I
PD W
55a .La
ww
IL
0 I
D-83
Figure 1: Setup for the Microwave Repair System
-MICROWAVE REPAIR- SYSTEM FORADVANCED COMPOSITES:
. H
A ......... D.C. Supply Source
B .......... 110 V A.C. Supply
C .......... Power-Supply.
D ......... tMlcrowave-Power Generator
E ....... Ferlt Isolator
F .......... Forward / Reflected Power Indicator
G ......... Tuner
H.......... Variable-Coupling Iris
I ......... Ap~licator
J ............ Material Under Process/Repair
D-84
CURING SYSTEMCONTROL COMPUTER
STACKED PREPREG
NORWVESNO
SINGNAL PROCESSOR
SINGLE O1 MUILTI-MODEMICROWAVE ItAV=T
D-8 5
Portable Aplicator,Large Area Joining
Advantages: * Large Area Processing
* Product Dissassembly
* High Safety Constraints~Push Handle
Microwave Sourcer0
Bar to /
Raise and
Lower Wheels
Aperture is similarto previousAperture Applicator .
Movable Safety MeshWheels
Safety Devices: Movable Wheels for good sheet contact
Surrounding Mesh Protector 0
Automatic Shut-off when lifted from surface
D-86 0
Taper App1licator:Strip Joining
Advantages: e High Field Concentratione Fast Heating Rate
I * Easy Material Handling
Taper Applicator with Supporting Equipment
Mg etrn claetof TaerAppiao
Lm
Waveguide
D-87
Microwave 0iav
Generator icuao Transfrmer
4-Power Control MR Sensor
SheetSI LAminattS n ft
Conveyer Belt / " 0
Travelling-Wave Applicator
0
Microwave Power To Water Load 0
Micrownvc Henting 2nd Joiing/Scaling of' Sheets ILmninates
SM&iar Skin
0
D-88 0
20
D-8
0
0
0
0
0
0
0
0
D-900
S
S
S
S
0
S
S
S
S
S D-91
p
pAPPENDIX E
HIGH STRAIN ACTUATORSI
I
S
I
I
I
I
I
HIGH STRAIN ACTUATORS
E-1
SOLID STATE ACTUATORS
TYPE OF ACTUATOR NATURE OF THE PHENOMENATRANSDUCTION EXPLOITED.,
Shape Memory Alloys Tiermo-Mechanical FerroelaSticPhase Change.
Piezoelectric Electro-Mechanical FerroelectricityOpto-Electro-Mechanic in a poled
ceramic orpolymer.
Electrostriction Electro- Mechanical IncipientFerroelectricityin a spin glass.
Contracti'e Polymers Chemo-Mechanical Differs inElectr o-Mechanical different types.Thermo-MechanicalOpto-Mechanical
M uscle Electro-Mechanical Action potentialpulses.
E-2
POLAFRIZATION CHNGE MECHANISMS USED TO CONTROLELECTROSTRICTWIE STRAIN.
(a) Homogeneous polarization of a Parselectric Phase
E-0O Pin ./2- PczE.
Stable zero strain state.
*(b) Induced polarization in a Relaxer ferroelectric
E& -~ ru vP P atE wLrf4rt1
= Stable zero strain state.
* (c) Induced Polarization change in Poled Ceramic
-oeE=OFwR nd-= PindaC
Unstable zero strain state (Aging change Of P)
(d) Micro -Macrodomain Poled Relamor
E- P0 jG P usE HystereticIStable zero strain state
v I: (e) Field forced Phase Change
E- 0 P= 0A4~ 0 P us E Double HystereticStable or unstable depending on system
(f) Field Forced Domain Change
E -0 Pin - 0 A2 4O P usE Hysteretic
Unstable zero strain state
E-3
THINKING BEHIND THE STUDY
The useable phenomena in the electrical control of the dimensions of an
irsulating solid are piezoelectriiy and electrostriction.
Xij = SijkJXd + bm ijPm + QrnnijPmPn
where xij are the components of the strain tensorXkj are components of the stress tensor 0
PmPn are components of electric polarizationSijkl the elastic compliancesbrmij the polarization related piezoelectric constants
Qmnij the polarization related electrostriction constants
1) No knQw solid has high enough values of b:ij, Qmnij to permit the inductionby external field of PmPn values sufficient to generate 1 % strain.
2) In ferroelectric crystals values of spontaneous polarization Ps are large
enough to give spontaneous strains up to 15%.
3) We must therefore eithera. explore phase chang which take us from a non-polar into a ferroelectricstate P = 0 to P = Ps;b. explore domain changes in ferroelectric crystals which reorient thespontaneous strain.
4) From equation I if PmPn are homogneous throughout the solid large valuesof strain xij may be induced at ZERO STRESS. Thus even in brittle solids, aswith thermal expansion, large induced strains are theoretically possiblewithout breaking the solid.
E-4
'I 1CS OF STUDY A'!'PENN STATEi\AVIICI(IALS RESEARCH LABORATORY
ie~ld Fl)Iced ,;%vilciijiiig fzoiii Anti feci 'Iclecuric to F-cr-roclccric I'hwcss ill
Lead Landimm Z ii Ihcolate'itazatc Statilatc Ceramuics
* Field FuincCLI Mictuduiiahl to Macrodoaii Switcinig ill Comipositions ofI Xcad La11i:Ii.iaiuIII Ziz-COMaelCitana-t Close to theC rlpllot-opiC HIsM
F ield Fuuce IParaclectric to IPerioclecti Switching in Single Crystal BariumTi lana IC
* )OInlaill Switcliiig inl BdrIO 3 Singic CrystalsSingle Field 90u Domain SwitchingOit hogoiial Field 900 Domain Switching
.;witclaiiig of'I.amiellar Wecdge Domains ill U110i3 C'ytas t) loduceUltr-a-high SitzaiL liinoiphi Configurationl
1 E-5
ANTIFRROELECTRIC to
EERROELECTRLC SWITCHING.
E--6
TABLE 1: COMPOSITIONS AND REFERENCE NUMBERS
No Composition
4 Pb.97La02(Zr66i%9Sn.)O 3
5 Pb.97La.02(Zr.5T 12Sn.3$)O3
6 Pb., 7La0,2(Zr.6Ti. 1 Sn.2)O 3
7 Pb.97La.02(Zr.6T.I 05Sn.23)O 3
8 Pb 97L 02(Zr.6Ti..Sn27)O3
9 Pb9,7 a.(Zr.,Tt1 1 Sn.2)O 3
10 Pb9,7La0,2(Zr,Ti. ,1 Sn.27)03
11 Pb.97La02(Zr6TiOSn.2)O 3
12 Pb.97La02 (Zrwri.,Sn.3)03
13 Pb.97La.02(Z7ji 1T', 2,Sn '0) 3
14 Pb9,7La0, 2(Zr,ri.1 Sn.3,)0 3
B-
400
4k4
00
E-8
w.
WaW
W
V 4 w.01
CLa
E-9
Table I1: Switching Data for Different Compositions 0
(a) Group 1 Compositions
No P(uC/cm 2) EA.F (Kv/cm) Ea (Kv/cm) X (ind)
5 30 30 35 0.18%
13 28 30 43 0.45%
10 36 28 60 0.5%
9 36 24 60 0.59%
7 36 22 58 0.52%
6 40 21 46 0.87%
(b) Group 2 Compositions
No P (uC/cm 2 ) E. F (Kv/cm) Eap (Kv/cm) X I(Ind)
14 31 44 56 0.35%
12 32 49 59 0.42%
8 30.5 52 68 0.37%
4 43 50 75 0.55%
11 33 45 60 0.45%
E-1O
Forwud switchin 9 in composition #5
10
9 (a)
8
4
.25 .5 .75 I 1.25 1.5 1.75 2
time (ps
40.0
(b)
30.0 . a U
20.0-U
10.0-
10.00 .0 , I , I , i I ! , , , I i i I ,
30.00 35.00 40.00 45.00 50.00
E (kv/cm)
(a) A sct of switching current -time curves under various applied fields(b) charges obsorbed per unit sample area due to switching as afunction of applied field for Pb0.97Lao.o2(Zr0.66Ti0.09$n0.5)O 3ceramic samples.
E-1.
3.0[ A
2.0 a. Ew53 kYcmb. E-65 kY/c
z c. E-76 k/cu
a0.0
-0.5 *
-2 a 2 4 6TIME (uSec.)
Backward Switching in composition #4. Field dependence
2.5 __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
__2.0
50 60 70 80
E (kY/cm)
E-12 0
II
I
aWiU
0U aI *
* U 00 0I
U 0U
F.,' U, --0 0 a W* U 0 _ 4~l63 U S C.)-L U
,E.
U 0 0
.0 0 0S - I z-m
U 0 =C.,_ 0 -
a -
U)I*SC a
'a
* II 0
U 0
0* 0 Q 9 0 0
- - S
NOLLV~UV1Od %
I
* E-13
MICRODOMAIN to
MACRODOMAIN SWITCHING.
E-14
PbZr0 3 MOLE % PbZr0 3 PbTi0 3
100 90 80 70 60 50 40 30 20 0 0
* FETET-J 10 AFE/J i i
30 1
(a)
E-15
0
E SCALE: 5KV/C4
e~j 0
CL
2SC 30 OC
34 C 40 CC
SO0 C 70 CC 100 OC
E-16
E MvICH E (KV/[H)Ia -5 u 5 10-10 -5 0 5 Iu
2~4U 4U
CC
~n.4
70
IOU
E-17
(A) )
I 0
zI z
-0.5 ,PLZT 7.5/53
LPLZT 7.5/65/350
0.0 - ! , .... 000 0. 0 .0.00 0.06 0.12 0.18 POLARIZATION SQUARE P 2(C2.M 40)POLARIZATION SQUARE P2 (C 2 /M 4 ) O
o 00
suaefpolazt P/PLT 8.8/65/35
position 4A ro tepaue. I •
U-).1 X0 0.0 0.2000.00 0.04 0.06 0.12 0.0 . 2 0.0M4
POLARIZATION SQUARE p2 (C 2 /,M4 ) POLARIZATION SQUARE P C/
1210..
PLZT 8/70/300C. I-PLZT 18/70/30 -
V101 & ._0.00 0.04 0.08 0.120.00 0.04 0.08 0.12 PLRZTO QAEP2( /
POAIAINSUR 2 (~M)POLARIZATION SQUARE P2 (C2 /M 4 )0
(A) transverse strain, s,,, and (B) longitudinal strain, S, , vssquare of polarization, P, in 7.5/65/35, 8.8/65/35, and 8/70/30 com-positions at room temperature.
Table 5: Field induced strain and related dielecmc damfor conventionally sintered PL.ZT ceramics
comp. r( 0C) Km K25 Ec(kv/cm) Pr(O x t(103) x1 (1O"3) X*xt
8/67/33 99 12000 5500 2.6 21 .81 2.5 .1
8/65/35 • 106 11350 4600 3.6 20 .82 2.3 .32
8/63/37 114 11300 4500 4.7 21 .76 1.9 .32765/35 140 15000 3000 4.5 28.4 1 3.1 .7
7/62../37.5' 160 16000 2900 5 27.2 1.2 3.7 .64
7/60/4 172 17000 3000 6.3 26.2 1.2 3.8 .4
7/58/42 180 17300 2600 8 22 1.1 3.2 .397/5&/ 190 17200 00 10 22 .94 2.3 .46/62/38 196 19000 2100 5 31 1.45 4.1 .58
6/6W0* 204 1800 2000 5.6 29.5 1.35 4.7 .576/58/42 7.45 29 1.32 3.9 .53
5/6&40 230 190 0 6.52 32 , .79 4.2 .535/58.5/41-'? 6.41 34 1.24 4.5 .59
5/56/44 8.5 12.1 1.6 A4/57/430 7.47 1.26 3.0
4/55/45 10 29.5 1.21 2.9 .55
'NMB compositos Tm: Temp. of dielectric mw umx Trasvem Sati ih=ed at 15kv/cm lqn: Maximm dielecric conSUM
xl: Loanitudinl snindcewd at 15kv/cm K25: Dielecmc consmnt a22C
x : Tnsvemr reimnue M aM
p E-19
HOT PRESSED PUT 7/60/32
OZ F CLEAR SMRACE
LUJ
0.6
0
0~I.2
10, 10, 10, 10 10, 10, 10, 10,
CYCLES
E-20
1.0
7/68/ Hot Uwm -. - 1--"h-.2' 7117 II11.3 -
- 0.0
0.4
C,,
CL .2
a.0
C-. 0.6
"' 10 2 1O ' 1 10a4 ]a 5 10 6 i0O7 10 9 10 9
CYCLES
E-21
CL1
((a)
0.0
10 10 10 5 0 10 *0
E-22
-4vU~lT 0 zi 44
ELICIRIC NILO UCYlUU
Low temperaLure 1)olarizaUoti and strainR cycles In a 9.5165/35 PUTF.F rctucuv.y 10 11&. Toulpmaurc -132'C. Cyclig field 30 ICVcsu.
E-21
0o
a)CC
C?~.
h- 0z 'D N
0 0 =U
ct 00CC~J Now 0)
Cct0& a:
ii~i~ w U)'
CX CMCOO IhmI~N
I'% 8 LL.
(c.) 0U) Cc
Z cc
E-24
CL co0 C)
0. C)
U, EU0)
CY z
U) 0)
o LU
0
z4w L
.
cco> co
0
LU C)
* E-25
Q)
C. NCL 4
Soo
0
@40o a Sam
0 ccMo.~~ ~ ~ h. L.c -
CJ ) -~a
0 0) Ca 0)
CC 0ow"
L 0 0
am~SME amm OC OC O 0G Cc
>0
0Uamac
0.0
E-26
xE
SS
in 0
.0 ccWu Cu
r- V
a)
OR.C C)00I U) -
a cc
- 0 .0
e0 IL0 CD
0 10 Cc :
Co -ECc* z W- cw V
ul E.-E5), 0 --o C13
~.1 ~ zwU
ch-
4
z I
E-27
o u 0 UL - 0
U9) U- a) 4DU) )U)mIjO Ln in a 0 - * 0 m~
0
CO.
0) 0)D
0~~~ 'Ln 0n I
r9
00
0~~~~ A% E Sf 0 0 L
>c.CC CL0 -
Ri
QCV C) CDZO 0n C
XLIz0-
E-2
U)M 0
U) z E)C
cnm
a)~toCl) * 0~a)QS
M cVj)
Cl
E--
.0
> 1 CO
0~I In E C......
C-4 0 = ,,.= o
> r,
cc 0
C0 Cl C a
0 11
cc~ > 0 0
- 0J".2j£ I' zC ....
,o w Lu. >
E-3
S_ cc
"aO 4)~ .2 r~r .. ~ 0N. x t-
&I rE-30 C
C-
(A) ULIM warn
* a
0Cc 0
z lo 0.0 .
0 -- 0
Lon cc 100h0L.ww) auw 0 0o 0 C 0 Ccho0~~ ~ _A SOUN 024)c.
IMMMM b.t U)4) OGMM. 0 O> 0 0~ cc~ 000M
0 3m... cc L.C. LCu0 o ~.~MW~ 0 C.) W
0 "i mOM w%@dMM VOWED0 0b~o ~0.50 .0 Cu 0C. 0 CO Qi-j Iw4 Iia
m0 . 0 ow0w 0 cm3cc
0 .
CD)
00 N
00
0
CMOW 0.00 C-
00
U/ m.- 00,80
0 00on- 0 . CL
weo 00
>m. C4 0
L) somms.0
o. "oo' GUo oi i .-
0)) LMsoa.L
O4 = * 0...e
0
E-32
*0 am0
00
0 c 0
oa 0 0 0 c0 U C zC. 0.C0 m 0
E E10cc 0 &M
00 0 a
.C 0 CL
Eo 0 OImE0 0 LM(/) CD 0 V
E 00 "E, m.)c
Cu 0 >
0u L 0Emwgg E 0 CA CA
0-U) LID Lm m t
_~~ 0 u E zCO > Cc 0 0" e
0 Cuo 0o.2.
cm1e 0
9 3L E-33
LIMITATIONS OF HIGH DRIVE
PIEZOELECTRIC AND EL CTROSTRICTVE
MATERIALS 0
GENE H. HAERTLING
CLEMSON UNIVERSITY0
E-34
PROPERTIES AND ASSOCIATED PHENOMENAOF PLZT FERROELECTRIC CERAMICS
PROPERTIES PHENOMENA
'vDIELECTRIC PHOTOCONDUCTIVE
PIEZOELECTRIC PHOTOVOLTAIC
PYROELECTRIC PHOTOCHROMIC
FERROELECTRIC PHOTOREFRACTIVE
ELECTROSTRICTIVE PHOTOSTRICTIVE
FERROELASTIC PHOTO-ASSISTEDDOMAIN SWITCHING
ELECTROOPTIC ELECTROOPTICSCATTERING
v/OPTICAL SURFACEDEFORMATION
v/ELECTRICAL RESISTIVITY SPACE CHARGEEFFECTS
* IsyiOY properi't0d~ ~ cI - -propertt-5
qf pha yuwwona
E-35
FACTORS LIMITING THE PERFORMANCE OFPIEZOELECTRIC AND ELECTROSTRICTIVE MATERIALS
1. COMPOSITION (ELEMENTS, PURITY, DOPANTS, ETC.)
2. MICROSTRUCTURE (DENSITY, GRAIN SIZE, PHASES)
-< 3. PROPERTIES (VALUE, VOLTAGE DEPENDENCE)
4. TEMPERATURE (Tc vs.OPERATING TEMP., SELF HEATING)
5. HYSTERESIS (ELECTRICAL SWITCHING, DOMAIN WALL MOBILITY)
6. AGEING (TIME DEPENDENT PROPERTY CHANGES)
7. FATIGUE (OPERATIONAL DEGRADATION, CYCLIC EFFECTS) 0
8. CREEP (CYCLIC MECHANICAL INSTABILITY)
9. ELECTRICAL BREAKDOWN (MAXIMUM VOLTAGE CAPABILITY)
10. SPEED (ACTIVATIONIRELAXATION PROCESSES)
11. SIZE (RESONANT/NON-RESONANT, FREQ. DAMPED EFFECTS)
12. STRESS (TENSILE, COMPRESSIVE, HYDROSTATIC, MISMATCH)
13. MECHANICAL STRENGTH (INTRINSIC, FLAWS, SURFACE FINISH)
14. FRACTURE TOUGHNESS (CRACK PROPAGATION, PRESTRESS) 0
15. ELECTRODES (TYPE, OHMIC CONTACTS, ELECT. MIGRATION)
16. STRUCTURES (MATERIALS, BONDING AGENTS, ENCAPSULANTS)
17. ENVIRONMENT (HERMITICITY, HUMIDITY, VACUUM, G-LOADING)
18. THERMAL HISTORY (MEMORY EFFECTS, POLED CONDITION)
E-36 0
:YP!CAL RANGES OF PROPERTIES FOR ?IEZOELECTRIC AND ELECTROSTRICTVEMATERIALS
Mater:al K Tan delta ;p r.3 d3 Q1 Q12
% PC/N (x10-16m 4/C 2 )
PZT 300-2000 .1-2.0 .2-.70 .3-.70 100-600
BT 500-3000 .2-1.0 .2-.40 .3-.50 100-300
PLZT 500-6000 2.0-8.0 .3-.72 .3-.80 150-700 .. ..
PMN 10,000-25,000 2.0-8.0 ...-- .020 -. 007
PN 200-600 .8-1.5 .1-.20 .4-.50 85 -- --
PZT 4 1300 .40 .58 .70 289
1ZT 8 1000 .30 .51 .68 225
PZT 5A .700 1.5 .60 .71 374
PZT 5H 3400 2.0 .65 .75 593 -- --
PLZT 7/65/35 1850 1.8 .62 .76 400 .022 -.012
PLZT 8/65/35 3400 3.0 .65 .78 682 .018 -.008
?LZT 9/65/35 5700 6.0 -- -- -- .020 -. 010
PBZT 73/27 8000 7.0 '-- .00 -. 040
E-37
LIMITATING VALUES FOR SELECTED FACTORS AFFECTING THEPERFORMANCE OF FIEZOELECTRICS AND ELECTROSTRICTORS
COMPOSITIONAL RE.RODUCIBILi 7: -0.01o
SHRINKAGE REPRODUCIBILITY: -0.1%
PROPERTY REPRODUCIBILITY: . 2%
USE TEMPERATURE:
PIEZOELECTRIC - T < 1/2 TcELECTROSTRICTOR - T > Tt (Tt = T of stable colarization)TYPICAL - T < -'CO degrees centigrade
SPEED: 0BULK- > 1 usecTHIN FILM - > 1 nsec
STRAIN HYSTERESIS: >1%
AC OPERATING FIELD: <15 kV/cm
DC OPERATING FIELD: < 40 kV/cm
ELECTRIC FIELD BREAKDOWN:BULK- 100 kV/cm (250 V/mil)THIN FILM - 500 kV/cm (1250 V/mil)
MAXIMUM LOAD UNDER STRESS: VARIES (gins. to kgms.)
AGEING: -2%/DECADE OF TIME 0
MECHANICAL STRENGTH:
TENSILE - 70 MPa (10,000 psi) (ac, - 4 K5LFLEXURE- 103 MPa (15,000 psi)COMPRESSIVE- 350 MPa (40,000 psi)REPRODUCIBIUTY - +20 %
E-38
SMAR'CERAMICSj
R.E. NEWNHAM-r r im A _ I L A
Y .A! . . .N /o* . ,.-- . m ,
.-m' -," -PA ... * . ,. . --. '~ .
7# .,,i k4 ':A!
STATE*COLLEGE
~; ' * P141LADELPI4IA
OFFICE OF NAVAL RESEARCH
E-39
E LECTR OMECHANwICALTRAN SDUCERS
ITNAAN
LINEAR N 0NLNA
PIlEZoELEC1piCry EECTRO STR I C~fO
DOMAIN CHAI4Ge'PUECf4EPHASF.Ci4AMG
sT I .9 U
Hsr- ~~8~T~
E-4 0
t$MART COMPOSITE COMPONENTS
C ge~m tC A L
-7 op~rCAL
o; - V 4f ANOCO rApsr'
PC~ EMpERATURE
P, 151L i E I
_PLC,
A PE
,Mww
*q. uvDANSDL)CPEijc
t< E
LtN
E-41c ~
PARTICLE~ 9 KPS IN nr osIt At)
14 A"MtX 0Ov3 NATR (X 3
P~RRGLS~ PEIRPORAM6 PERF RAIl
~~RA3- -.M
L~t4GIT1)9(We CetmmC 943
Hot4sycoto 3.1#4Y~ *
3.3 VYCO
I S. * F
0.
2*29E-42
CONTROLLED COMPLIRNCE
SEN SOR $STRESS
* SENSOR DETECTSINCREAS(NG 5TRZESS
,REUCE.S HEC-NT 'OF101 -7 =i=L , ROtRAC ,q CTU)A 70i-,<
* COK51NAJT(ON (4IMLCSULTRA5OFT SOLID
E-4 3
-- m lows.
TOYO-T A * ELECTRONIC,IODULATED*SUSPENSI ON$~
rsM~s -P!VNPtE Z E L. SCFR IC
SENSOR ON PISTON ROoF' sc ~o.R ~l
Ro A1D e LE.CTRO N I C CONTRO
UQGSNESS iq LG'N VO LTAGES
, .. ." ...
I s tit" A ENLARC T 2r1is
F KA!CHANGlING,~~,K
TOSo F' 5 VN 4~rvL(PCATQ(4.) IN
E-44
,SMART SHOCK ABSORBER
_- PISTON ROD
""i"."RO::AD SURFACE
SENSOR
PIEZOELECTRIC
;A. .IFEC AT(ON135wIAR.y LF L C, %,k,/",:ANM NEL
SWlrfCNrz VALVC
H.TSUKA, 7 NAKOMO AND Y YOIKOY4?,69E, tE WORKSHOP) 1990,
E-45
kCTIVE VIBRATION CONTROLACTIVE SUSPENSIONS
*RA1LfRoAT)C% oMo)
*, AUTMOBILES (NAGAI)FLEXIBLE STRUCTURES
ROBOTIC ARMS (FUIKUDA)*11 • I ... - ' - : ," ,*w ,(
*SACIE STRUCo.TURES (KA-SNIwASE)** CANTILEVERED PIPES (DOKI)
VIBRATION & NOISE CONTROL
*, F~rGIG N.AMSRS. (=TANAKA)
* VIADUCTS & BR,\DGE5 (YAHAGI)
* ACTIVE BEAR IN S f(NONAMI)
REFERENCES IN .3S.NlE. 19B.7 ONE-46
~ACTUATOR DESIGNS
* i snobLT L-E1 TAC!.G FORCE
M A SMALL D1SPLACEMEA
*o ~J-0.H 1 , DM.000. M ETA L SMALL FORCE'' Mlmft%.. SH ImA H- BIG DISPLACEME'NT,
, Moo jf .DI.K.
INTERMEDIATE
FCRCs. S DISPLAC.E-
E-47
PIEZOELECTRIC I4ATRIKPOLtD
DIREC~T EFFECTC -ttC
LAR1ATON 4o o ooo )
_______E E FECTS
~$ d. 5.0 00 0
l 0
0
E-4 8
CRESCEN~T m SHAPED 2%OmZH;;PROPwONE (MOONIE)
SHAPED MtETALELECTROES ;?N POLYWtE R
0
0
0
ES.A1AC t t t tT
ooY ,.i C4VIPLA
INTERNAL 'TeNSILe STRSSSAC'IN6O PeL2oceAmicIN RADIAL 1v-%,rok
E-4 9
frCAVITY SIZE EFFECT.*HYDROSTA~tC PIEZOEL.ECTRIC
CHARMGE CEFCEi
* oR~ G o ~~Pum
ISIAALLCAN!Y
0 1000
HY(DROSTATIC PRESSU (PST)
E-50
........%- -o - - . .zMOONIE ACTUATOR
LVDT ,~V/p7
GLAS S __N(
R ASS
AL LA 6i' A
DISPLACEKEN1 fz
02
0 2co 46o0.8~ &W a"YOLA GE
E-51
D.LECTROSTRICTIVE MOONIEACTlUATOR
^110
~-,4r(,BASS Et III CARS5OUDER. FOIL 13ONWIINGNO POLCNG RSQutREV
ISPLACE'MEN4T5o LA
EQU IVA LENTL33 : 600 C
VOLTAGE (V)
SUSAWARA, Q~C.)(U, PENN STATEE-520
MULTIMOONIE ACTUATORIt ALL ACTUATORS 1-2 rnrr THICK
PMN MULTIMOONIE PMNE 0 MOON IE.
2 @VsoAi {PZT9MOOMIE
=:4
PMN DISK~PZT DI SK
200 400 too 1000
VO LTAGE (v)
S. SUGUWARA (PENN STATE -1 SOPHIA)
E-53
PIEZOELECTRIC COEFFICIENTS_
EL.ECTRN
PIEZOELECTfRIC- CHARGE CEFCE
* 4 r.' =0 lef 15 CTR O td~ 1.
VORMAL- PIEZoLLECTRIC (-3 3 I 5 p/
TUNWABLE PISZOELETRICS(MOONI~oS6 RELAXORS)
E-540
Lu w
z
Iq Z
z w2Li. 0 c
cc oC
us C, ( 0 l(3
wE-IL
00
0
zl z
oL c j p0u Uj 0k ft-3- -
C) L0
0 .
00 C,0
C,) 0 m'(Cl
L1) CC, -ClCL
0tL <v
cri C Ln'( m WOW
c 0 0 u
E-560
.~z
0 In 0
w U)
U.
LU>k. LU >)/
LU
() 0LLUC
* U) 0D
0 LU~U).iJ
Znm 0 -JD j
LU .U.I~ U)... uEIEhmhhhhi LL L
M L ( U =WO
O U) t
F- ) ) U) U
C 0 0U
E-57
zLU
La.
Lui
00IL-
0
CO -
E-58
U)n
La N\'U
0 cmc> +> a
z 3
Iin
uOn
--. o oE-59
w0
0
N in 0
mlJ
0 I
II
U. UU
.jI
Ina C1
Mt/P II Md 3 aaV
E-60
~cc
0 F0
00
0Im 0 J
0~
j m~
E-I--
NX
00) E
E6
LUO
C.)
w wO
0mm 0
(n W i z1-. 0
0
m
E-62
wUoD m c -
E
o E6
> 6n
w C.) U
UU -JLU z1.oz 0 C)QmU 0 a
w. w U z
( U.
IE16
zI-C,,
0 Li0
area >.Oz
Z 0z M
00 c0
0 0U
E-4P
Q0
Mv C~4
LU
LU. a
o C ( ,cc. 0z z 0
o U.z m
U.U
E-65
oD aY 00w
In v-
Ln
-ILU0 ..
CC um U~0U 4n
I-. z Luo U m.0z 0.
4t Lu
mmmlc5
*mmm6
x
C) > LL
(,)0
-Jow
C.)
cn Z u
LUu
LUU
um m J
wz
E- 6 7
CD 00
0 In-
Cm.))
'JLL- LU
LU z)0 *l 0
m LU
IIq >0 conU CO z z0
Za z 0LU M LL
Pca z um 4LU L LUum
Cmm
ca 0
1 -2
Va 0 go0
K aa
a v~ cu
me20u Z
00
a
E-680
o *,Zcc 0
tn I
C3)
00
Ij.1 w -J
__Uz 0
0z m
zL i- 2.
Lu i z z~W EL
CAU. ~ -
LUQo
LU omCO-
0 4K OA i@ c
uswU
r-6
0z0
Ch
oc 00(4.) I- -
w 1-0
LU
LU
co j COmoW
oA z w C
o) -
U m '.... 01.
00
u
L. a JU
0.
zo
E-70
Nl Cl) X
~~jLJ
Ln
0
cn
z
LU E.J
1mm. Z 0
0 .0
LU LL
~ ( LU z 5 0
2 t ZU. Z
LUL 4U.m2.2
-
00II
0
zCl)-- z w
-) ' 0 2 0Wu W M. ZM
P CLaWLZ lw
L- 2
_s 0ww(I) US 2 .JIA
-~ - LOP- Z . Z U m
IZ u
Z Zwc
lU Z F
LU WC3 U)
LL.L
(4.) 0
crS
WI-zca ow
E-720
U, U1uj
00
0 00
LUU-j
CIL z
20-U )zm L
0
LU ~ L -
LL
LU
(I)(I)
E-73
N o Cl)
a cocc C14 0) 0r
0Y)
.j.00LU
-I) U)
o j
Z~ U)
LU CD
0 z
E-74
2
0
0
0 0
zC.V iULLU
0 LL >zF
0
z LU
LU< (
+ L-I
rc0NLU
LU
E-75
00
o 0* uj 0iV~ N =
I- 0
00
cn-I z
'UI>I
'u Iml m
z cz
0 U C) m
'U z 0-z 0
LA-a - 00 0 wU zu.Z
U.E
CcoCLI
CC--I
00
E-76
w Cl) mxI-0w wZ
C~D0N
LU e0
LIU
w U 0
Lp, L LU
zM -U)U
z CCz 0 OW01m LU U. L
Lu zmmm, 2
U.2
C00.0Z0
INI
E-77
41 Im
c c c
,~ U.o0cn.Cmn.C.C
E-78
z0Cl,
LuJ
crOz
zui > LUI-j
o 00
cm
zaz
oJ 0oUl
z 0
LU
E-79
V -V NOLL03S
olS
E-80
C
CAC
'Jo Ct
LUU
CLC
* 4m
uj Q a
LU
z
o- < -z n J l L LJLL
LU M .M-
2u = - 'uLf_ C6 j '-
Ql ~Cl)LuJ Z L Z
2 Cl .. .. cau n 40UJCc
LU Uu 0 w .0 U M
0 C) lh
M- -. M<
U~Z Lu PI).(Lu L - Lu0 .J 0 Q.
=(*h LU Zo
C.) 0 u 1 (4IJ21 0z 0
PC- ~ ~ mW: 0 LQ LILU CflcZ r1.0-- C.) L
t.Am >mcu a0j0;:Qc
.j CO~ LL. U.j~ 0 -ft 4 Llw amm L
*ZL 9 0W--~> Mo c 3)'2
TM aLL.Z m C%
Us8 4 Y 40 U > U l
APPENDIX F
IMPROVED PZT THROUGH CONTROLLED POWDER
SYNTHESIS AND BULK POWDER PROCESSING
U)wz a)z C
Cl) obw
00 -J
0U)~
.C.0 z
* 00 Cl)
O~m *m
C/Cu
-1
how
.0
o0. -0a
'0
o .o .
>- o
0. 00 so
E &,0-
ES
•0 .0- amsm 0 a
igoo
&M) 0N0. 0.a - CL-
owCV 0E0ma).Q-a) 01-0 0.
0mE
Cc-
0 2 r. 0 a
0-~ E 0. 24... 0 a)
C.) 0 I00)1
Lm~
0~ N.z CDE Eo. 00a
4) a)V 0.00
0-0
F-2
0 0
0 0
Cl) 0cn cc
r E 0o 0 cn 0o ha 0
0.C 0) >%4) M 4"
Z o 0o 0 0 2D CLw E0 w 00 0-0
z h 0 0 I 0.M Ch0. CcE 00 E -. 0
0 0 CA =woD 0.w E 0 A-w0) 0gc m~ a 00&
0. cE 0 0 01
0-c0.LUc
w0
00
_ 0
mE
2IC) NIt!I @).. -4
0 cm tm c
G)G)LU)~a * *D
F-3
050
cC Cl U) 0)
0fl ea m
a) L
one 0 CL aCL 0
0 0 E
Omm =
CL C
0- E -w E5 N
0 am am
w 0 am -E
~U 0 0 a)
:o '0 0 ie cz 0su~a~a
o0 0.0 CU LOU
"no ~ a C OUNA N 0 N 0 0 m
cl) U) 3: U) C cc0cl' now NC C LL. 0)
E= 00 EELuMI LED
C, cc ~ 3: mLuCL) mLa) a -5m E*
p m amn am
~LL z~
0 NEW
F-4
_) CU
aL -- a
- -0 0)
0 &- 0 :C
C C
) N . 00 cc ow) ?. a0
E Ca) Cc. C ~ C
0. E Ea CU 0~c0O~A- a0 N c
0. aE L4Z1 00 E ow) r c
0 0 c
0 a 0
CLwam Q
mJm
IL
0-
00
z 0L0 0
4 0.0z 0
0
C) N
0. _Cr) CL
w, 0.
c CDLu 0
0.
5 0 0O CL 0CCj DCL CL
mI ~ -- C. Co UAm 0. 0
z .06- EC, C) 00N0- jo
o. 3o 0m 0.>AE. N CMVE ) 0 0C)0 0 00) 0 0.0
a) a-a a- 0
a- Am -CL c C
0 0
F-6
EG 00
CUC 0 0
-6-
- w ) cc ... 0
0 N0
c c o a m
r Z-E 0 .0
>0 ~A 0
0 0.' ;o.
E Em
0 0
0
m E4
F-7
C/ wq 0
w 0w
L.J 0 N
zcc4)ma>6 0Z i Z
LM L.
a:L C0C) 0
LL.M0
LU)
I a.
F-8
E IQ
N 0n CNr)c
a=W
OLD 0
-0 0 0 (Q 0 a
00
LL)
W 14t 00
3:C) Ir) C0 0. C
CC.
<(00
I-c,,
Om _E)w ocaOz 0
U(U
ca)0 >t CD
0 E
L0 L. 0
am- 0
0~0w A A2:
F-10
C/4)
LM 0
0M LM
0 c0
0 E
_J 0E (U(U) - - >
CM 0CC Lu 4)LM 0
00 -M
unO 0 40 Ow)
0. 0
an 0 N ) 0A 00
- C-
F-li
COC
LLI 0
N cn_M 0
0 E
00
Ow-
0 > 0020)
0 EM0 J 0
w~ ca 0
E E W w0 00
CL cc
A A A A AoC
F-12
.JQ0
0
z 4m)
-E4-o
0 C\JN o E 6
N 0 E
C)C
0 d
<0 0L L0. 0 a aEm - 0 coG
0w 0t 0 E d0M Lm 0
a.o 000l
0 0
00
-F 1
E8LU _ 0~ 0) a) 0
(Iv o C aL
0 E0-~E 0 -o0 Oa )
cc0 0 0 0 0 ()0) 0
S 00)
0 0~ior '7o: (OD 'A>- ECLo0
ui000 0 0 Oc0)n
00
o cA00
00
0 1 0
F-14
0~ -n~j
*~ ~ (Y C\ Y0)~ (Y) c)~-~I C
C\cN M (
Zi 0
4)
00
qCf ))' 0 ~-
CLcE
I- 0~IJU~ huwC C4:-
F-15C~q
COO
Din~C!) LC)
Wa CO0 c
0)
0 C) C
L
0i 000
04) LOO . 00 a
LLO. LO 0
4))
o Lu
Cu CO)
0)
0
AA A
F-16
w
C\ I"--VI-\E LL V)
0 o o)4-
CL Cl) E)
uj cu
0 Ul
0 () CO
E IpLL 0 0 N --
00£4C) 4-,W M,
Z :-~ D 0
w 5
0
F-17
CCO
N 0
>=e CLuj 0
4.)
LLW
.rL 0 L
L CLT AA A A
00LM 0
ow-1Z
LUU(~~ -OZ~
C.)
uJ N.) N%.10 0
pjO M
1 0 1
F-19
(u
- StU
- 0t
I!
0F 2
ow
0o o
)UEm
omom
C 0. 4-brA E
L] - co I
Q~ Zemwl
w 0 C Z
p ~ IF.21
Sf
Ia
LoS
CL r
Sc
I 44
I47.
14 e
00
399
0mt- *
0 0
Immola
Cc
EoE
a-
000
0
o N
F-25
0%0
00
Eoc
6 z5
0F-6
Cc C
04..Im-
0~ C)
0l C E)CtG) tI
00Z
CC 00
ci
UJ >
F-27
rill,
LM0
C-
0.
DC)~ C)I
I rarr
,G)~ 004).) .fI>
w '__ _ _o
N (A
ec -
cr
C.)F-28
l) C
) cc1*
-1 IM L
0 Ii;
mw 0
I I II II I I I
wrm m en
u CC
Q- -(J)~ - -
.0F-29
CCV)
00
Cl)... (u ______________
0 0 ix
0C. C> 0f t0 0_
_ 0I
0F 3
, I
4) O
* I4
Iw c)I
0 0
00 L~0
- m - Ip',I
Qii
-1
cn
F-3
CC)
cc eqt 00nC)
-C
T
F-3
O )
cc__I__ __E~1m.
(A Cu . I-m__~-4.- cv~cu
0 1 0
00 0
C.C)
0) tn 0 n Q0 0 0 0 04
o620 0 0
* r o-
00
E N- C-4 0 F
0 V- 00
tp -, \-C-
0 EI 2 I 0 i4C.,w 0E
U- '%L ~C%4
0 N
00 10 0C0
Do 000 94m
D(*) E 0
F-34
00 ~~ 0000
(1 0 0
CF)(80 cc ONO
hD C0 I I
_C4 0 N qr ON4G) -00
a) 2 00 0... r- - - C 00( ) ) W." r ~ ~ t
S! -
IM IDo 0I-. %.- 2 ,e-e
,0) C0*0 cc -n t-- - -- -
Em 0 m~t 0n N N
EI 1 1 I 4
U) V-4
-0
-Z - r
00
> >~ -~ E
-- - - --N
S F-35
0 _m
Z 0
Cc 0.blow
oWoSEAS
0 PC
00>k.
00
-0
6 z_
woWo
Z w
F-36
IC
00ow
CC oI.
F-37
Actuator comparisons to Multi-layer Ceramic Capacitors
Area ML Capacitors ML Actuators
Dielectric BTBi PZTPMN
Electrodes Pd, Pd/Ag Platinum
Design 40L/32A 500L/125A
Cer Thick 10-35 p-m 150 p-m
Overall sizes < 3mm cubes 6 x 150 mm
Casting 50p, water 200 p-m, org
Printing Thick Film Same, 2X
Lamination 2 mm pad 25 mm pads
Dicing Blades Saw
Burnout 0/4 hrs 60/168 hrs
Firing 900/13000C 10000C
Termination Silver/glass Diff Comp
Leads None Ins. Wire
Coating None * Si Varnish
Testing Elect & Vis & Mechan
Quantities 15 mil/day 100's/day 0
Cost Pennies 1 O's dollars
F-38 0A\ ?V
Percent Shrinkage.. " - -& -"
U, 0 N)4. n0 0 ; . Cm 00 0
00
0 *O
__ 0
CU MI CDp 0
* 0D .fQ
CL -7
Z 0
*0
TI-3
ZZZZZI ..
ve MLC
"Tfl-0
F-40
00
thFFQSE iecrieooG
F-41
Materials Systems Inc.
INTRODUCTION
Develop and manufacture advanced piezoelectricmaterials and devices for defense and commercialtransducer applications.
- Flexible Piezoelectric Composites- Piezoelectric Composite Actuators- Piezoelectric Actuator Assemblies- Composite Transducers for Medical Ultrasound
TECHNOLOGY
o Process technology for ceramics and polymers:- Extrusion- Injection molding
o Near net-shape forming of complex ceramic shapes.
o Piezoelectric materials and composites expertise.
o Materials technology transfer and device prototyp;ng.F
F-4 2
0~
04
INJECTION MOLDING PROCESS ROUTE
PZT Powder Particle Ball mill in closed containers.-Size Control (if needed)
Compound OgncBne(mix with wargncxine
Wax
Pleccle ranuatePlasticizer
I Surfactants
Burnout Controlled atmosphere.
IJ .. remove carbon andArFlash bind particles.
Sinter
[Fin ish Face off surfaces flat and parallel,apwply electrodes, package and ship.
F-440
.Materials Systems Inc.
PIEZOELECTRIC ACTUATORS
Observations
o Performance Characteristics:- Fast response- Large forces- High voltages- Small displacements
o Several routes to enhance displacement performance:- High-strain piezoelectric materials
* - Strain amplification using compound actuator designs- Multiple actuator assemblies
o Displacement needs may require a combined approach.- Compound actuator assemblies
o New high-strain materials and device designs are inthe development stage.
o Need to integrate materials, design and manufacturingfunctions to best meet manufacturing requirements:- Lab-stage materials- Lab-stage designs- Design with system goals, manufacturing viability
and cost-effectiveness in mind.
F-45
Materials Systems Inc.
ACTUATOR ASSEMBLIES
Focus: Actuators incorporating materials and designsat the limit of the current state-of-the-art: S
- Compound actuator assemblies
Examples: - Flextensional strain amplifiers- Piezoelectric ceramic/polymer composites
Challenges: Materials/process issues: 0
- High-strain piezoelectric ceramics- Ceramic/metal/polymer interfaces- Complex shape fabrication
Assembly requirements:
- Alignment- Joining- Packaging- Testing
F-46 0
.. Materials Systems Inc.
PIEZOELECTRIC CERAMIC/POLYMER COMPOSITES
1-3 CompositeElectro e Area
PZT Rod.
0
PolVmer
Production issues remaining to be solved:
- Handling large quantities of fibers.
- Composite assembly.
- Acceptable cost.
- Systems application and integration.
Manufacturing Approach:
-Use injection molding to form the ceramicelements to net-shape.
F-47
Materials Systems Inc.
PZT CERAMIC PREFORM FABRICATION
CERAMICPREFORM
REMOVABLESTOCK
,,_REMOVABLE
SINSERT
-TOOL BODY
INSERT DEFINESELEMENT SHAPEAND LAYOUT
PREFORM LAY-UPTO FORM LARGER
ARRAYS
F-48
Materials Systems Inc.
FLEXTENSIONAL STRAIN AMPLIFIERS
Void Brass
: Bond
PZT
Production issues remaining to be solved:
- Alignment during assembly (reproducibility)- Joining approach (device life)- Design of multiple actuator assemblies (ruggedness)- Acceptable cost- Systems application and integration
Manufacturing Approach:
- Assemble prototype quantities manually:(needs tooling for alignment during joiningand lay-up, and development of partdimensional specifications)*
- Full-scale manufacturing would be facilitatedby redesign to simplify joining and alignment
F-49
Materials Systems Inc.
SUMMARY
o Emerging actuator materials and designs hold promisefor considerably improving strain performance overconventional piezoelectric ceramics.
o Pilot scale manufacturing of flextensional actuatorsfor field evaluation is technically feasible now.
o Larger quantities would benefit from redesign forimproved manufacturability and operating life.
o Cost-effective manufacturing of advanced actuatordesigns may require net-shape forming of complexceramic shapes; injection molding is available for thispurpose.
0F-50