D. Parks, K. Sinding, S. Zhang and B. Tittmann
Penn State University, University Park, PA 16802
[1] Motivation
[2] Objective
[3] Work in Progress
[4] Results
[5] Conclusions
Motivation
 Ultrasonic techniques have been widely applied for non-
destructive evaluation (NDE) of material properties and
structural integrity.
 The ability to apply typical ultrasonic techniques in high
temperature environments is desired for the
 monitoring of process variables, such as viscosity of melts
 evaluation of structural/material integrity in harsh
environments, e.g.,



turbine blades,
combustion engines and
nuclear reactors.
 Aircraft engine turbine and fan disk cracks have led to
disk burst and catastrophic engine failures:
Problem Statement
 United Airlines 232, Sioux City, Iowa, (1989) – 111 fatalities
 Delta Air Lines 1288, Pensacola, Florida, (1996) – 2 fatalities
 Air Florida Airlines 2198 – takeoff aborted
Part of the DC-10's fuselage after the
crash.
Damaged Engine of N927DA
What are the Applications?
 Characterization of flaws is
an important goal of
ultrasonic NDE in aircraft
 structural health
monitoring to achieve
retirement for cause.





Aluminum wings/fuselage
Composite components
Turbines
Steam pressure pipes/vessels
Aircraft Disc brakes
Application: develop in-situ monitoring
techniques for a nuclear reactor environment
Core
Research reactors:
• Material specimens
• LE fuel specimens
Commercial reactors:
• critical structures
6
6
Fukushima Daiichi, Japan
7
Background
 High temperature piezoelectric materials have been
surveyed previously.
 For example, the commercial piezoelectric material
PZT possesses a Curie temperature of about 350o C but
has a maximum recommended operation temperature
of 150-250o C.
 As a result, buffer rods/ultrasonic guided waves have
been utilized to keep the piezoelectric materials out of
the high temperature region for extended periods of
time.
Limitations
 Limitations of a guided wave system include,




depending on the required waveguide geometry, the
following limitations:
Dispersive behavior limiting bandwidth
Sensors size
Cumulative attenuation losses over the waveguide
propagation length
Losses due to material discontinuities along the
ultrasonic propagation path
New Materials
 The limitations imposed by this methodology have
been one of many driving forces that are leading to the
development of new piezoelectric materials with
operational temperatures exceeding 350o C.
 The relatively new high temperature capable materials
include high Curie temperature ferroelectric materials
and non-ferroelectric single crystals.
 Additionally, depending on the intended operating
frequency, electrical resistivity becomes problematic
for high temperatures and low frequency operating
conditions
Candidate Piezoelectric Materials
 High Curie Temperature
 Few stable phase transitions
 High electromechanical
coupling
11
 No problematic atomic
species
 Reactor waste requirements
 Kaz
Show that AlN performs in γ
up to 18.7 MGy
AlN
 The group III-V material Aluminum Nitride (AlN)
belongs to the Wurtzite structure and point group
6mm and is not ferroelectric.
 No phase transitions exist aside from the melting
point at 2800o C
 resistivity of high quality crystalline samples is on the
order of 10 MΩcm at 200o C. The dielectric constant
has been measured to be 8.5 and the piezoelectric
strain constant has been estimated from thin films to
be roughly 5 pC/N.
Transducer Design
 Reactor waste requires relatively
short half life isotopes (Al 6061
primary material)
 Aluminum foil coupling is used
after heat treatment for a
Brazed contact
 Waveguide Al 6061 for in-situ
characterization
 Develop radiation tolerant BNC
cable
 Fused-Quartz insulation
 Aluminum conduit
sleeve/inner conductor
 50ft Low Loss RG 213 U
 Carbon-Carbon backing
 Monolithic bulk z-cut single
crystal AlN centered at 13.4
MHz
13
 Purpose: Understand Sensor
Test Setup
Behavior
 Does radiation reduce
piezoelectric effect tending
towards an isotropic state?
 Is there increase dielectric loss
due to defect generation and
ionization
 Is there mechanical failure due
to transducer design?
 Measurements
 Impedance (tanδ)
 Pulse-echo A-Scan through
aluminum 6061 waveguide
 In-Situ Test Time : 3 Months
 Fast Neutron: 1.85x10^18
n/cm^2
 Thermal Neutron: 5.8x10^18
n/cm^2
 Gamma:26.8 MGy
14
Results
Pulse-echo amplitude
measured in-situ
 Reactor-on status
measured by loss in
signal amplitude
 Pulse-echo amplitude
stable during off-state
 Amplitude relatively
stable 15% deviation
 Fluctuations need to be
isolated from changes in
transducer mechanics
15
Lithium Niobate
 Ferroelectric material with a Curie temperature in excess of




1000° C depending on the stoichiometry .
3m crystal symmetry with the corundum structure. The
36° rotated Y-cut is quite sensitive in the longitudinal mode
of vibration with a coupling coefficient of 0.48.
known to lose oxygen at elevated temperatures, particularly
at low oxygen partial pressure.
decomposes at 600° C even in oxygen at atmospheric
pressure.
oxygen loss is an activated process so for 170 hours before
observing their decomposition.
YCOB
 Oxyborate crystals with general formula ReCa4O(BO3)3 (Re
= rare earth element, abbreviated as ReCOB) used for
nonlinear optic applications.
 no phase transitions occur prior to their melting points,
those being on the order of 1400-1500° C.
 ultrahigh electrical resistivity at elevated temperatures. For
example, YCa4O(BO3)3 (YCOB) possesses a resistivity of
2×108 Ωcm at 800° .
 The dielectric permittivity, piezoelectric strain constant,
and electromechanical coupling factor of the (XYlw -15°
/45°) cut were found to be on the order of 11, 6.5 pC/N, and
0.12, respectively, with little variation in the range of room
temperature to 950o C
Test Fixture for High T
Normalized echo amplitude while at 550oC continuously
for 60 hours in an open tube furnace
Thermal Ratcheting Experiments
Pulse-echo Waveforms for YCOB
Sol-gel Spray-on Fabrication of BiTi transducers
 Fabrication method in a nut shell:
 Powders of ferroelectric materials are
added to aqueous solution of mixed oxide
precursors.

Powders increase viscosity such that sol-gel
can be sprayed with an air gun onto surface.
 Substrate is sprayed, then pyrolyzed.

Repeated until desired thickness is
achieved.
 Substrates can be metallic.
 Sinter, followed by electroding and
poling.
Sintering
 Typical sintering methods have proved to be
destructive and difficult to handle for the
sol-gel deposition method,



Induction sintering has proved most effective
Microwave sintering
Blow torch sintering
Corona Poling
 Corona Poling
 Poling method which does
not require electrodes
 Metallic substrate is
grounded and heated (150200o C), metallic needles
are brought within
proximity of sample surface
 High voltage ( ~10 kV)
applied to needles,
resulting electric field
aligns electric domains
within ferroelectric
material
Comb Transducer Fabrication
directly on Pipe
Laser ablation has been
effective at creating
an accurate electrode
pattern for launching
guided waves in pipes
Bi4Ti3O12 High Temperature Response

Amplitude
remained
relatively
constant until
625 oC…
followed by
sharp decline.
Ultrasonic response was lost
after 1000 oC
Bi4Ti3O12-LiNbO3
Pk-Pk Amplitude (V)
2
1.5
1
0.5
0
0
100
200
300
400 500 600
Temperature (C)
700
800
900 1000
OBSERVATIONS
 the long-term in-situ testing indicates that all four
materials are suitable for operation at 550° C for at
least 55 hours.
 carbon-carbon backing material is a limiting
component. A solution based on a porous Al backing
was found but thorough testing on this backing
remains to be completed. This backing material is
easily created by compressing Al foil at 550° C under
150 psi for 5 hours.
Fixture Details
Capacitance and Attenuation
OBSERVATION
 The YCOB crystal exhibited a much less pronounced
change in dielectric properties after heat treatment.
 It is expected that YCOB is more stable at high
temperatures than LiNbO3 which is known to deplete
its oxygen particularly at low oxygen partial pressure.
 YCOB, AlN and LiNbO3 exhibit stability in ultrasonic
performance through the heat treatment of 950° C for
24 hours and 1000° C for 48 hours. Any variations
observed were less than the experimental error.
Summary
 High temperature piezoelectric crystals, including YCa4O(BO3)3,
LiNbO3 and AlN, have been studied for use in ultrasonic transducers
under continuous operation for 55 hours at 550° C.
 Additionally, thermal ratcheting tests were performed on the
transducers by subjecting the crystals to heat treatments followed by
ultrasonic performance testing at room temperature and 500° C.
 .The changes due to the heat treatments where less than the statistical
spread obtained in repeated experiments and thus considered
negligible.
 Finally, in-situ measurements up to 950° C, of the pulse echo response
of YCa4O(BO3)3 were performed for the first time, showing stable
characteristics up to these high temperature.
 Sol-gel Bi4Ti3O12-LiNbO3 gave good performance to almost 1000
°C
CONCLUSION
 At atmospheric oxygen partial pressures, 48 hours of
exposure to 950° C or 24 hours exposure to 1000° C had
no significant effect on the efficiency of ultrasonic
transduction of LiNbO3, YCa4O(BO3)3 and AlN.
 The oxidation of AlN, known to occur at these
temperatures, had no significant effect on the
ultrasonic transduction efficiency, but the difficulty to
achieve high quality AlN single crystals limit their
applications greatly.
 YCOB crystal was found to be capable of efficient
ultrasonic transduction to about 1000° C.
OUTLOOCK:
From Navier’s equation, the temperature dependence of guided
wave propagation depends on the material properties.
 Lame
 G   u  G  u  
2
 u
2
t
2
34
Theoretical Dispersion Curves
 Effect of thermal expansion (cross-section)
 CTE ≈ 6 μ/K in Zircaloy
 295K (70K) vs. 700K (800F)
 Δthickness = 0.015 mm
 Δwidth = 0.05 mm
 Negligible at 150 kHz
35
Theoretical Dispersion Curves
 Zircaloy bar specimen:
 λshear=19.7 mm @ 150 KHz
 Bar width = 20.625 mm
 Cannot use Rayleigh-Lamb equations for bar!
 Solve via SAFE technique
36
2D Semi-Analytical Finite Element (SAFE)
 Define analytical solution in the wave propagation
direction
 Orthogonal function
 Discretize cross-section; Boundary conditions
 Eigenvalues of system = wavenumbers
 Wavenumbers → dispersion curves
Cp 

k
Cg 

k
37
Group Velocity (SAFE)
38
Group Velocity vs. Temperature (150kHz)
39
Thank you for your attention
 Any Questions?