Space Qualified NDE

1
Space Qualified NDE HM Technology
December 6, 2005
NextGen Aeronautics Dr. Shiv Joshi University of
South Carolina Dr. Victor Giurgiutiu
2
Presentation Outline

• Current projects
• Objectives and goals for space application
• Applications and innovations
• Commercialization
• Technical approaches and challenges
• Structural health monitoring (SHM) with
  piezoelectric active wafer sensors (PWAS)
• Impedance acquisition module miniaturization
• PWAS durability and survivability
• SHM demonstration on complicated space panel
• Composite materials
• Thick structures
• EUSR and phased array advances

3
Structural Health Monitoring (SHM)

• Passive SHM records flight parameters, loads,
  strain, environment, vibrations, impacts,
  acoustic emission from cracks, etc.
• Active SHM detects damage, cracks, disbonds,
  delaminations, etc. (embedded ultrasonic NDE)
• Research Aim Develop embedded NDE sensors for
  active SHM

(Giurgiutiu, V. Zagrai, A. N. Bao, J.
Piezoelectric Wafer Embedded Active Sensors for
Aging Aircraft Structural Health Monitoring,
Structural Health Monitoring An International
Journal, Sage Pub., Vol. 1, No. 1, July 2002, pp.
41-61 )
4
Current Health Monitoring Projects
5
Space NDE HM Goal
Assurance of Structural Integrity for Space
applications with minimal weight and cost impact
6
Objectives

• Development of resilient space qualified
  non-destructive evaluation (NDE) and
  health-monitoring technologies for on-orbit
  inspection and maintenance of aerospace systems
• Validate sensor technologies through space
  qualification tests
• Validate weight/cost impact using selected
  application
• Identify certification barriers and risk
  mitigation strategies for commercialization

Strategy built on concepts developed during the
last decade in smart materials and structures
area, especially by PIs from NextGen and the
Southern Carolina University (SCU)
7
Applications

• Potential Applications
• Adhesives, sealants, bearings, coatings, glasses,
  alloys, laminates, monolithics, material blends,
  wire insulating materials, weldments
• Thermal protection systems
• Complex composite and hybrid structural systems
• Low density and high temperature materials
• Aging wiring

Near-term Applications Applications targeted for further development
Detection of cracks, corrosion and disbonds Detection of cracks under bolts Real time and in situ monitoring Micro-meteor impact damage assessment Damage characterization for durability and life prediction Planetary entry aeroshell validation Electronic system/ wiring integrity assessment Early detection of damage
8
Innovative Concepts

• There are three distinct innovative concepts
  incorporated in the our technology.

• Excitation of preferential Lamb/Rayleigh wave
  modes,
• Utilization of phased array concepts,
• Utilization of software algorithms rather than
  hardware for beam forming and signal analysis.

These concepts make it possible to develop an
ultra light, small footprint, low energy
consuming, reliable and low cost structural
health monitoring system.
9
Space HM Overview
Phase II Development Testing
Integration Flight Testing
Certification Marketing
Commercialization

• Development of PWAS NDE methods on complex
  structure
• Development of phased array(reduces quantity of
  sensors)

• NDE HM system for specific structural
  application
• Space certification testing
• Validate weight/cost savings

TRL 3
TRL 8
10
Path to Commercialization

• Develop NDE HM system using PWAS technology for
  a selected space application
• Validate PWAS technology by space certification
  tests
• Validate weight/cost impact using selected
  application
• Identify certification barriers and recommend
  risk mitigation strategies for commercialization

Satellite
Space Station
ReusableLaunch Vehicles
Shuttle
11
Technical Approach
SPACE HM Phase I

• Space Qualification Testing
• Coupon Tests for Design Environment
• Subcomponent Testing under Loads
• Design curves for minimal sensor placement

Space Application Results
Phase II
X-33

• System Integration Test
• Lockheed Martins X-34composite LO2 tank

composite LH2 tank with integrated TPS
12
Technical Challenges
Challenge The reflection/scattering from a crack
is superposed on reflections/ scattering from
geometric irregularities, bonding and mechanical
fasteners. This results in noisy
signal. Solution It could be resolved by using
differential signal method. In this approach, the
signal from the pristine structure is subtracted
from the signal received in the present state of
the structure. Challenge Assessment of the
severity of damage. Solution Near-field
Damage Correlation Coefficient Deviation (CCD)
damage matrix Far-field Damage Probabilistic
Neural Networks (PNN)
13
Other Issues
Durability USC performed durability and
survivability tests. Thermal cycle loading to
check the interface between PWAS and structure.
Exposure to outdoor environment (rain, snow,
etc). Submersion in various liquids (salt water,
hydraulic fluid). Various adhesives, coatings,
and wire combinations were tested. Strain USC
performed PWAS response of structure under
mechanical static and fatigue loading. Hardware U
se wireless sensors and miniature IC
chips. Software Signal conditioning, continue
development of EUSR, Correlation Coefficient
Deviation (CCD) damage matrix, Probabilistic
Neural Networking (PNN) Curvature USC performed
tests on panels with various curvature. Composites
Composite panel tested in Phase I.
14
Conventional Ultrasonic Methods
Pulse-echo signals
Pulse-echo
Pitch-catch (acousto-ultrasonics)
15
Piezoelectric Wafer Active Sensors (PWAS)
Conventional ultrasonic NDE transducer
PWAS array
Lamb wave
P-wave
16
How are PWAS Used in SHM?
Propagating Lamb waves
Standing Lamb waves (E/M Impedance)
17
PWAS for SHM State of the Art

• Chang et al. (Stanford)
• Inman et al. (Virginia Tech)
• Yuan et al. (NC State)
• Cesnik et al. (Michigan)
• Adams et al. (Purdue)
• Kessler Spearing (MIT)
• Cawley, Soutis, Culshaw, et al. (UK Imperial
  College, Sheffield)
• Boller et al. (Germany EDAS ? UK Sheffield)
• Balageas et al. (France ONERA, CNAM, INSA, )
• Galea, Ye et al. (Australia)
• Giurgiutiu et al. (South Carolina)

18
Overview of Test Panels
19
NASA Spacecraft Panel 1
Panel 1
20
Pitch-Catch Detection Panel 1
21
Disbond Detection with Pitch-Catch Method
Pitch-catch signals
Damage index
22
P-E Instrumentation Layout Panel 1
23
Pulse-Echo Detection Panel 1
24
Disbond Detection with Pulse-Echo Method
Signal received at PWAS a7 has echo from the
disbond DB2. Signals for PWAS a8 a20 are
pristine
25
EMI Instrumentation LayoutPanel 1
26
E/M ImpedancePanel 1
27
Disbond Detection with E/M Impedance
Resonant frequencies spectrum showing increased
amplitude for the signal received at the sensor
located on the top of disbond DB1 (PWAS a2)
28
E/M ImpedancePanel 1
29
Corrosion Detection with E/M Impedance
Resonant frequencies showing shifted peaks for
corroded area CR1 (PWAS b30) vs. undamaged area
(PWAS b31)
30
E/M ImpedancePanel 1
31
Crack Detection with E/M Impedance
Resonant frequencies of the sensor close to the
crack CK1 (PWAS a30) and the sensors in a
pristine area (PWAS b29 and b34)
32
CB1 Detection with Pitch-Catch Method
Pitch-catch signals
Cracked
Pristine
PWAS pair b16 and b15, b18 and b17 pitch-catch
signals are pristine pitch-catch signals.
33
CB1 Detection with Pulse-Echo Method
Pulse-echo signals
34
CB1 Detection with E/M Impedance Method
E/M impedance spectrums
35
Impedance Miniaturization Objectives

• Design and develop a miniaturized, field portable
  impedance analyzer

The overview of the entire development of the
proposed miniaturized impedance analyzer
36
Proof-of-Concept System Architecture
A miniaturized E/M impedance analyzer will have 3
modules
Impedance measurement circuit

• Reference signal generation
• Voltage (V) and current (I) measurements
• Digital signal processing for calculating the
  amplitudes and initial phases of the voltage and
  current
• Integration method
• Correlation method
• Discrete Fourier Transform (DFT) method

Signal Generator
Desktop with DAQ card
Zoom in circuit
To DAQ card channel 1
To DAQ card channel 2
Calibrated resistor
From function generator
Free PWAS
37
Integration Method
The input signal S is multiplied by a sine signal
and a cosine signal respectively, and then the
results are integrated over a time duration of T
( T is the period of the signal S).
38
Correlation Method
?Definition of Cross-correlation function
Consider two signals of the form
,
Then,
39
Discrete Fourier Transform Method
Consider signal
Discrete Fourier transform (DFT) of x(t)
Where q the number of cycles of signal x, N is
the number of samples.
40
Experimental Results Real part
Comparison of measurement of real part of
impedance spectrum of a free PWAS with different
methods
41
Impedance Spectrums for Disbond Detection

• Real part impedance spectrums of PWAS a1, a2 a3
  measured by HP4194A impedance analyzer.
• Real part impedance spectrums of PWAS a1, a2 a3
  measured by low-cost impedance analyzer using
  freq. swept signal source.
• Impedance spectrums from PWAS a1 and a3 located
  on area with good bond are almost identical.
• The impedance spectrum fro PWAS a2 located on the
  disbond DB1 is very different showing new strong
  resonant peaks.
• Both of the low-cost impedance analyzer and
  HP4194a impedance analyzer can detect the
  presence of disbond DB1.

42
Objective

• Explore the durability and survivability issues
  on PWAS associated with various environmental
  conditions and fatigue
• Improve properties, layer deteriorates in time
  under environmental attacks (temperature,
  humidity, etc.).
• Improve properties, layer deteriorates in time
  under fatigue effects

PWAS-structure bond layer
43
PWAS Durability under Thermal Cycling
44
Damage Index

• Development of suitable damage metrics and damage
  identification algorithms
• The damage index is a scalar quantity that serves
  as a metric of the damage present in the
  structure.
• RMSD

45
Bonded PWAS Impedance Spectrum under Outdoor
Exposure

• Settling in effect.
• Significant change has been recorded.
• Damage index shows the impedance changes.

46
PWAS submersion tests

• Distilled water
• Saline solution
• Hydraulic fluid MIL-PRF- 83282 Synthetic
  hydrocarbon
• Hydraulic fluid MIL-PRF- 87257 Synthetic
  hydrocarbon
• Hydraulic fluid MIL-PRF- 5606 Mineral
• Aircraft lube oil MIL-PRF-7808L Grade 3 Turbine
  engine synthetic
• Aviation kerosene
• RESULTS 60 weeks without failure except in
  saline solution which failed after 15 weeks

47
PWAS Impedance Spectrum Under Submersion Exposure

• A little impedance changes in distilled water
• The PWAS submerged in saline solution survived
  only a little over 85 days due to the detachment
  of the soldered connection
• The corrosive effect of the saline solution.

48
Large-Strain PWAS Testing
PWAS

• Minimal changes up to 4000 me (0.84Y)
• Failure at 7300 me (1.13Y)

Failed PWAS _at_ 7300 ??
49
PWAS Fatigue Survivability Tests
PWAS
Stress concentration
50
Composite Plate

• Material A534/AF252 Uni Tape density 1.54 g/cm2
  
• 16-ply quasi-isotropic composite plate
  (0/45/90/-45)2S
• Average axial Young modulus of the plate E
  206GPa
• 2.25 mm thickness 1236 mm 1236 mm size

90o
0o
51
Square PWAS Experiments

• 7-mm square PWAS, 0.2-mm thick
• Rectangular pattern
• 250 mm distance
• Parallel and perpendicular to surface fiber
  direction

90o
0o
52
Fiber Direction and Frequency Effects On
Pitch-Catch Transmission

• Fiber direction 0o a) S1 transmitter, S2
  receiver b)S1 receiver, S2 transmitter.
• Fiber direction 45o a) S1 transmitter, S3
  receiver b)S1 receiver, S3 transmitter.
• Fiber direction 90o a) S3 transmitter, S2
  receiver b)S3 receiver, S2 transmitter.
• Fiber direction 90o tune burst amplitude is the
  maximum at every frequency.
• Fiber direction 0º tone burst amplitude is at
  its minimum.
• Maximum difference between received amplitude at
  90º and the other directions is at the lower
  frequencies.

53
Data Collection For Thicker Structures With
Complex Geometries
54
Tuning for Thick Structures (Old Method)
55
Tuning for Thick Structures (New Methods)
240 kHz
56
Phased Array Technology

• Phased array physical characteristics
• Typically now a 50 mm long array of 10 square
  PWAS
• th 0.15 mm
• Extremely light weight

57
Motivation Phased Array Technique

• Phased array concentrates the waves in certain
  direction and scans target range
• Scanning without mechanical movement
• High inspection speed
• Flexible data processing capability
• Improved resolution
• PWAS phased array techniques
• Delay-and-sum beamforming (Johnson and Dudgeon
  1993)
• Using guided waves for structural health
  monitoring
• Using Piezoelectric Wafer Active Sensor (PWAS) to
  generate/receive Lamb waves

58
Generic Beam Forming Formulation

• Delay-and-sum beamforming
• If a propagating signal is present in an arrays
  aperture, the sensors outputs, delayed by
  appropriate amounts and added together, reinforce
  the signal with respect to noise and waves
  propagating in different directions.

Reflector
Array
f1(t)
?1
f2(t)
?2
z(t)
?
fm(t) input signal from mth element ?m
delay for mth element signal wm weight for
mth element signal z(t) total output signals
of M element
.
fm(t)
? m
59
Geometry

• Define Origin phase center
• Parameters vectors
• Near- and far-field
• Near field use exact traveling wave paths
• Far field assumption

Far
Near
60
Wavefront at a Reflector P (r, ?0)
Near field
Far field
61
Optimization Ratio of d/r

• Ratio d/r is used to show the effect of spacing
  d.
• It indicates if a reflector is located in near-
  or far-field.
• It can be expressed as
• D - the span of the array, D(M-1)d
• Conclusion the directionality is best in far
  field with good mainlobe to sidelobe ratio.

M 8, d/? 0.5
62
Optimization Number of Elements M
Comparison of the beamforming of M 8 and M
16 in far field at 45º

• Conclusion Better directivity for M 16. The
  more elements the array has, the finer the
  mainlobe will be with smaller sidelobes
• Disadvantages limited available installation
  space for large size array and wiring complexity

d/? 0.5, d/r 1/(M-1)5 (far field) at ? 45º
63
Optimization Steering Angle ?0

• Effective steering range for 1D linear array is
  0º180º. When directed to certain ?0, beamforming
  differs from angle to angle.

• Conclusions
• Beamforming is symmetric about 90º
• Within 0º90º range, beamforming gets worse when
  ?0 becomes smaller
• With ?0 increases, directionality is improved
  with suppressed sidelobes
• The array has an angle below which the
  beamforming is bad. The critical angle becomes
  smaller with larger M

64
Future Work Mini-Array Design

• Optimization option
• smaller d/r
• Mini-array using scaled down elements
• If the size of PWAS shrinks to half, the size of
  the array D will decrease by 2M
• Therefore, ratio of d/r decreases to half since
• Frequency tuning the high frequency requirement
• Sweet triple point

Wave mode plot for Aluminum-2024-Ts, 3-mm thick
65
Mini-Array Experiment

• Broadside hole detection
• Broadside hole 80-mm away from the arrays
• For comparison, using both 5-mm (mini-array) and
  7-mm (regular) PWAS array
• Conclusion mini-array can detect the damage
  correctly and clearly

7-mm square
66
SC Work Sample
Tensile specimen used in Crack Growth Imaging
Experiment
67
Fatigue Testing with PWAS(Durability and
Survivability)

• VG-1 was scanned at a frequency of 372 kHz
  during 110,000 cycles with a cyclic load of 400
  lbf to 4000 lbf
• No signal could be seen without filter
• Signal could be seen with filter

68
ASCU Filter Demonstration
69
Crack Growth Imaging

• The data shown in these images was taken while
  the specimen was cycling by the system to be
  delivered to the AF
• A filter developed the LAMSS team can be used to
  remove cycling interference
• 800 lbs to 8k lbs (r 0.1)
• 5 Hz cycling
• Note Echo received from the slit does not
  overpower echo received from the crack

70
Crack Growth Imaging
Baseline-30 mm-data taken while under cyclic
loading
71
Crack Growth Imaging
22k cycles-35 mm-data taken while under cyclic
loading
72
Crack Growth Imaging
30k cycles-40 mm-data taken while under cyclic
loading
73
Crack Growth Imaging
42k cycles-50 mm-data taken while under cyclic
loading
74
Crack Growth Imaging
48k cycles-55 mm-data taken while under cyclic
loading
75
Crack Growth Imaging
58k cycles-60 mm-data taken while under cyclic
loading
76
Future Work

• Perform sensor layout tests
• Adapt wireless sensors
• Miniaturize hardware
• Develop damage detection software
• Develop health monitoring software
• Perform durability and survivability tests
• Evaluate weight and cost Impact
• Perform structural subcomponent test
• Perform structural component test
• Plan technology transition

77

• QUESTIONS