Post on 11-Jul-2020
205 Portland St • Boston, MA 02114 • 617.447.2172 • http://www.metisdesign.com
Seth S. Kessler, Ph.D. President/CEO
Metis Design Corporation 15 July 2014
structural health monitoring
multi-functional materials
lean enterprise solutions
PZT & CNT based SHM Systems
for Impact Detection & Localization
In-Space Inspection Workshop 2 of 24
• Metis Design developed & validated system through SBIR funding
AF03-T017 – Intelligent SHM Infrastructure (hardware)
AF06-097 – Adaptive Damage Detection (software)
USS Independence (N10-T042)
Triton UAS (N12-125)
Blackhawk (N12-T007, COST-A & others)
• System focus on low mass, low power, expendability, retro-fit
novel sensor/algorithm design for large-area coverage
distributed intelligence on a digital sensor bus
multifunctional capabilities at each node location
MD7-Pro Digital SHM System
© 2014 Metis Design Corporation
In-Space Inspection Workshop 3 of 24
MD7-Pro Structural Sonar
• Analog sensor base for impact/damage detection
• Greatly reduces typically required sensor density
• 1 PZT actuator & 6 PZT sensors in small package
• Facilitates both active/passive beamforming
Vector
Locator
nodes
Damage
40 mm
© 2014 Metis Design Corporation
15 mm
In-Space Inspection Workshop 4 of 24
MD7-Pro Acquisition Node
• Digital node for distributed acquisition & local computation (15 g mass)
• Greatly reduces mass of cables & centralized hardware, eliminates EMI
• Facilitates both active (guided wave) & passive (AE) detection methods
• 8 breakout analog & digital channels + built-in triaxial accelerometer & temp © 2014 Metis Design Corporation
50 mm
7 mm 5 mm
40 mm
In-Space Inspection Workshop 5 of 24
MD7-Pro Acquisition Node + Structural Sonar
© 2014 Metis Design Corporation
In-Space Inspection Workshop 6 of 24
MD7-Pro Low Speed Channel Validation
0.00
0.50
1.00
1.50
2.00
2.50
-3000 -2000 -1000 0 1000 2000 3000
MD
7 M
eas
ure
me
nt
(V)
Strain (µ)
Strain Gauge Calibration & Validation
MD7080
MD7081
0.00
0.50
1.00
1.50
2.00
2.50
0 5 10 15 20
MD
7 M
eas
ure
me
nt
(V)
Number of Fractured Traces
Crack Gauge Calibration & Validation
MD7080
MD7081
Model
0.00
0.50
1.00
1.50
2.00
2.50
0 20 40 60 80 100 120
MD
7 M
eas
ure
me
nt
(V)
Temperature (C)
Thermocouple Calibration & Validation
MD7080
MD7081
0.00
8.00
16.00
24.00
32.00
40.00
48.00
56.00
64.00
0 20 40 60 80 100 120
MD
7 M
eas
ure
me
nt
(k)
Temperature (C)
Thermistor Calibration & Validation
MD7080
MD7081
© 2014 Metis Design Corporation
In-Space Inspection Workshop 7 of 24
MD7-Pro Accumulation Node
• Digital node for 64 GB data accumulation & global processing (20 g mass)
• Can support up to 100 Acquisition nodes on serial bus
• Hosts complex C++ embedded algorithms w/FPGA & 2 GB RAM
• Gigabit Ethernet + USB access to data, programmable interface © 2014 Metis Design Corporation
60 mm
7 mm
40 mm
In-Space Inspection Workshop 8 of 24
Data Analysis & Reconstruction
© 2014 Metis Design Corporation
+
Each node processes phase-coherent, location independent “sonar-scan”
Sum scans incoherently to form composite image
Logic imposed to compensate for view area obstacles
color represents # of standard deviations above mean of damage-free data
In-Space Inspection Workshop 9 of 24
• Single MD7 node detection on 2mm thick Al plate with 20 rivets
36 impact events of ~20 J of energy from falling 1 cm semi-spherical mass
half of impact on each side of rivet line
• Hybrid passive/active detection demonstrated
36 passive/active auto-triggered measurement following impact events
6 manually triggered active measurements with a fastener removed
36 manually triggered active measurements without any impact or damage
Performance Evaluation
© 2014 Metis Design Corporation
In-Space Inspection Workshop 10 of 24
• Results collapsed to a single scatter plot of raw localization
prediction by re-centering all impacts to a common origin
100% detection (36/36) following impact events
no false triggers recorded at pre-set threshold levels
mean error for AE localization ~ 25 mm
predictions cluster relatively closely near origin relative to size of plate
no trend observed for results obtained on one side of fastener line vs other
Passive Mode Impact Detection Results
© 2014 Metis Design Corporation
mm
In-Space Inspection Workshop 11 of 24
• Results collapsed to a single scatter plot of raw localization
prediction by re-centering all impacts to a common origin
100% detection (36/36) of ~0.5 mm deep dents following AE detection
no false positives indicated (0/36) following non-impact scans
mean error for GW localization ~ 50 mm
more scattered than AE, but predictions still group relatively close to origin
no trend observed for results obtained on one side of fastener line vs other
Active Mode Impact Detection Results
© 2014 Metis Design Corporation
mm
In-Space Inspection Workshop 12 of 24
• Results collapsed to a single scatter plot of raw localization
prediction by re-centering all impacts to a common origin
100% detection (6/6) of hand-tightened fasteners
no false positives indicated (0/36) following non-loosened scans
mean error for GW localization ~ 5 mm
least amount of scatter due to massive local stiffness change
essentially translates to localization within 1 fastener position
Active Mode Fastener Detection Results
© 2014 Metis Design Corporation
mm
In-Space Inspection Workshop 13 of 24
• Conformal assemblies for composite & metallic host structures
central carbon nanotube (CNT) layer is core to these properties
surrounded by electrically insulating layers (film adhesive and/or GFRP)
selective electrodes integrated to steer current flow
• Little impact to physical structure, 100 - 200 µm & 5 - 10 g/m2
can be co-cured with composite laminate
can be installed over composite or metallic skin in secondary process
• Enable multi-functional capabilities: anti-icing, health monitoring
Conformal Multi-functional Assemblies
© 2014 Metis Design Corporation
In-Space Inspection Workshop 14 of 24
• SHM improves reliability, safety & readiness @ reduced costs
sensors add weight, power consumption & computational bandwidth
cables add weight, complexity, as well as durability & EMI concerns
scaling SHM for large-area coverage has presented challenges
• Advantages of proposed CNT-based SHM methodology
CNT “sensors” can actually improve specific strength/stiffness of structure
can use thinner/lighter electrodes such as metal-mesh or direct-write
simple to scale over large structure, maintains good local resolution
Structural Health Monitoring (SHM)
© 2014 Metis Design Corporation
Traditional SHM CNT-based SHM
In-Space Inspection Workshop 15 of 24
• Damage shifts CNT links in affected zone, increases resistivity
nearly linear increase in % resistance change with impact energy
< 1% change in resistance away from impact zone
• Surface & sub-surface images produced in post-processing
20 joule impact caused ~10-20% resistance change (no visible change)
40 joule impact caused ~20-30% resistance change (no visible change)
60 joule impact caused ~40-60% resistance change (no visible change)
Fine Grid Impact Results (AF08-T023)
Flex
circuit
CNT
sample
Silver electrodes
© 2014 Metis Design Corporation
In-Space Inspection Workshop 16 of 24
• Detection sensitivity strong function of CNT network aspect ratio
2400 mm² CNT w/160 mm² damage yields ~25% in resistance increase
same damage in 1 m long strip of same width would yield ~2% change
10 mm² damage would still be over noise floor
• Simple 2D network resistor model in good agreement with data
Sparse Electrode Notch-Cutting Tests (N111-067)
© 2014 Metis Design Corporation
In-Space Inspection Workshop 17 of 24
• Impacts below threshold of 30 J had <0.25% change in resistance
impacted surfaces exhibited >1% change in resistance after 30 J impacts
majority of specimens showed increase of ~15% after 110 J impacts
possible to increase CNT monitoring patch length to 1 m with 0.1% change
• Variability due to impact events, could be observed in “dents” too
Sparse Electrode Impact Results (N111-067)
© 2014 Metis Design Corporation
In-Space Inspection Workshop 18 of 24
Impact Test Acoustic Response (N111-067)
© 2014 Metis Design Corporation
7 J Impact 108 J Impact
7 J Impact Zoomed-In 108 J Impact Zoomed-In
In-Space Inspection Workshop 19 of 24
• Resistance is proportional to strain for low displacement
load/displacement curves for all specimens are in close agreement
tensile-side resistance increases due to CNT network being stretched-out
compressive-side resistance decreases due to CNT being pushed together
• Permanent resistance increase after 25 mm deflection (>400 N)
Sparse Electrode 4-Point Bent Results Under Load
© 2014 Metis Design Corporation
In-Space Inspection Workshop 20 of 24
Unloaded Bend Test Results (N111-067)
© 2014 Metis Design Corporation
In-Space Inspection Workshop 21 of 24
1m CFRP Submarine Propeller Test Specimen
© 2014 Metis Design Corporation
In-Space Inspection Workshop 22 of 24
4-Point Bend Results: Loaded vs Unloaded Results
© 2014 Metis Design Corporation
• Same trends observed in 1 meter specimen as smaller coupons
tensile-side resistance increases linearly with enforced displacement
compressive-side resistance decreases linearly with enforced displacement
• Permanent resistance increase after 25 mm deflection (>4 kN)
In-Space Inspection Workshop 23 of 24
Effect of CNT on Laminate Mechanical Properties
© 2014 Metis Design Corporation
• 4 sets of ASTM tests performed
professionally by testing house
• CNT surface layer statistically has
no effect on any mechanical
stiffness or strength properties in
normal operating strain ranges
Izod
Specimens
Average
Strength
Baseline 33 ft-lbs/in
CNT on surface 37 ft-lbs/in
Impact ASTM-D256-10
Tensile ASTM-D638-10
Compressive ASTM-D695-8
In-Space Inspection Workshop 24 of 24
Technical & Business Contact
Seth S. Kessler, Ph.D. President/CEO Metis Design Corporation
617-447-2172 x203 617-308-6743 (cell) skessler@metisdesign.com
© 2014 Metis Design Corporation