fabrication of a multi-physics integral structural ... · ¾direct-write (dw) technology can be...

Fabrication of a Multi-Physics Integral Structural

Diagnostic System Utilizing Nano-Engineered MaterialsDiagnostic System Utilizing Nano-Engineered Materials

Seth Kessler, Ajay Raghavan and Christopher Dunnj y g pMetis Design Corporation

Sunny Wicks, Roberto Guzman deVilloria and Brian Wardle

10 Canal Park • Cambridge, MA 02141 • 617.661.5616 • http://www.metisdesign.com

Massachusetts Institute of Technology

Advanced Composites & CNTs

© 2010 Metis Design Corporation 2010 PHM Society Conference 2

Carbon Nano-Tube (CNT) laminates are a natural progression for aerospace composites due to their superior specific strength & stiffness

Present CNT Challenges

• Several issues have hindered effectiveness of CNT engineeringdispersion: uniform distribution of CNTs in epoxy matrix for homogeneityadhesion: CNT-matrix interfacial bonding for effective load transferorientation: non-random alignment of CNTs to amplify propertiesextension: building macroscopic structures by linking individual CNTsextension: building macroscopic structures by linking individual CNTs

• Most investigators have focused on addressing 1-2 of these i i di id ll i ldi i l i tissues individually, yielding marginal improvements

• Simultaneously achieving all 4 necessary to achieve trueSimultaneously achieving all 4 necessary to achieve true potential of CNTs for enhanced composite structures


Aligned CNTs @ MIT

500 µm 100 µm 1 µm

• MIT has developed a novel patented CNT fabrication processCNTs grown radially aligned on fibers or on substrate to be transferredgood alignment dispersion adhesion & yields high CNT volume fractiongood alignment, dispersion, adhesion & yields high CNT volume fraction

• Atmospheric pressure chemical vapor deposition (CVD)self aligned morphology 1010 1011/cm2 of continuous CNTs (7 10 nm OD)self-aligned morphology 1010-1011/cm2 of continuous CNTs (7-10 nm OD)rapid forest growth of > 2 microns/second (up to 5 mm long)


MIT “Fuzzy Fiber” Laminates


Motivation For CNT-based SHM

• SHM improves reliability, safety & readiness @ reduced costssensors add weight, power consumption & computational bandwidthcables add weight, complexity, as well as durability & EMI concernsscaling SHM for large-area coverage has presented challenges

• Advantages of proposed CNT-based SHM methodologysensing elements actually improve specific strength/stiffness of structureconformal direct-write (DW) electrodes lighter & more durable than cableconformal direct write (DW) electrodes lighter & more durable than cablesimple to scale over large structure, maintains good local resolution


Traditional SHM CNT-based SHM

Direct-Write Technology

• Subject of Phase I NASA SBIR research, (PHM '09 paper)Direct-Write (DW) technology can be used to replace traditional cablessilver inks & copper sintering patterned on structure for conductorspolymers & ceramic deposited on structure for insulators

TRL f DW t h l i i t l t 8• TRL for DW technologies is at least 8DW implemented by Boeing on commercial aircraft production linesuccessful durability/longevity testing, FAA approved technologyy g y g pp gy

5 cmSHM nodes75 cm

5 cm


CAN/power link

CNT-Enhanced Composites for SHM

• Aligned CNTs greatly enhance composite laminate propertiesmechanically improve impact, delamination & fatigue resistancemultifunctional capabilities introduced by conductivity & piezoresistivity

• Present research exploits CNTs for in-situ damage imagingdirect-write (DW) electrode grids applied similar to LCD technologyin-plane & through-thickness resistance measurements collecteddamage breaks CNT links around affected zone, increases resistivityg ysurface & sub-surface damage images produced in post-processing

Multiplexing micro-switch

Positive electrode

CNT

columns


CNT-enhanced composite laminateGround electrode rows

Proof of Concept (POC) Approach

• FFRP laminate fabricatedalumina fiber satin-weave cloth dipped in solution of 50 mM iron nitrate catalyst coats fibers with CNT “seeds” inside tows of plyCNTs grow radially aligned 20-50 μm with modified thermal CVD method2 plies stacked by hand layup2 plies stacked by hand layupWest Systems™ Epoxy system infused for 12-hour RT cure~50% fiber volume fraction & ~ 2% CNT (115 x 25 x 2 mm)

• Silver-ink electrodes applied w/masked silk-screening process14 parallel conductors patterned in short-dimension on top plate surface14 parallel conductors patterned in short dimension on top plate surface4 parallel conductors patterned in long dimension on opposite surfacetraces were all 1.5 mm wide and spaced 3 mm edge-to-edge


POC Experimental Procedure

• Electrical resistance measured by multimeterinitial data showed significant variability due to contact resistancesubsequently thick wires were bonded to traces using silver epoxyresulted in < 1% change (< 0.1 Ω for ~10 Ω trace) across 10 repeat trials

• Centered impacted with 13 mm diameter steel ball at 75 ft-lbs calibrated to just initiate surface micro-cracking with trial impacts no damage visible to unaided eye but microscopy revealed crackingno damage visible to unaided eye, but microscopy revealed crackingpost-damage resistance measured between pairs & through thickness

3 mm 3 mm


3 mm 3 mm

POC Results

• In-plane resistance between parallel pairs of adjacent traces“short” traces averaged 9 Ω resistance pre-impact> 100% change in middle traces, < 10% change for outermost trace pairs minor changes < 1% measured between long trace pairsappears to be more sensitive to surface crackingappears to be more sensitive to surface cracking

• Through-thickness resistance at each virtual grid point56 through-thickness grid points averaged 20 Ω resistance pre-impact> 100% change in middle points, < 10% change for “left” half pointsresistance offset introduced by cracks across traces on “right” half pointsappears to be more sensitive to delaminationpp

In-Plane Resistance Change Through-Thickness Resistance Change


Further Validation Research

• Method validation with more comprehensive test matrix3 FFRP specimens (115 x 25 x 2 mm) observe effects of multiple progressive impact events 15-45 ft-lbs

• Process improvementsrefined silk screening process to reduce trace resistance & variabilityrefined silk-screening process to reduce trace resistance & variabilityflexible circuit frame for "overwrite" connection with DW traces8 x 32 traces 1.5 mm wide, all traces spaced by 1.5 mm (4x resolution)


Multiplexing Hardware

• Original data collected by hand with multimetersome variability introduced by manual connections & settling timesfewer trace combinations measured because of testing time ( ½ day)

• New multiplexing hardware designed dedicated to this research• New multiplexing hardware designed dedicated to this researchconnects directly to flex-frame via FFC cable, PC via RS-232dual multiplexer banks to select 2 traces for measurementconstant current applied through 2 traces, measure 16-bit voltageresistance directly related to V/I

• Allows many more combinations to be measured quickly1140 total measurements (including both sides of traces)t t ti d d t d t i d ttli ti (1 d t 1 h )test time dependant on determined settling time (1 second to 1 hour)


In-plane Resistance Changes

No isible damage as present in an of these cases• No visible damage was present in any of these casesnearly linear increase in % resistance change with impact energy< 1% change in resistance away from impact zone

• Appeared relatively localized to the actual impacted region15 ft-lbs impact caused ~10-20% changes30 ft lbs impact caused 20 30% changes30 ft-lbs impact caused ~20-30% changes45 ft-lbs impact caused ~40-60% changes


Through-Thickness Changes

Same trends obser ed for in plane s thro gh thickness res lts• Same trends observed for in-plane vs through-thickness resultswitness specimen testing indicated complete fracture at 50-60 ft-lbsno visible micro-cracking due to decrease in void fraction (2% to ~0.5%)

• Appeared to effect width in impact region relatively uniformly15 ft-lbs impact caused ~2-4% changes30 ft lbs impact caused 4 8% changes30 ft-lbs impact caused ~4-8% changes45 ft-lbs impact caused ~8-10% changes


Alternative Approaches

• Alternative dynamic piezoresistivity schemes being exploredacoustic emission (AE) measuring momentary resistance changesenhanced penetrating thermographic NDE with applied voltage

• Same hardware & flex frame can be used to measure dynamic i t h lf i d h tiresistance changes or self-induce heatinginitial “pencil-tap” experiments verifies that AE can be detectedinitial thermographs demonstrate method feasibility g p y

PROP

Impact between DW electrodes indicated

No impact indicated heating away from site

RIETA


ARY 100 mm100 mm

Future System Vision

• Sensing system integrated into vehicleCNTs introduced at ply level to be included in laminate manufacturingflex-frames built in standard sizes, installed at component levelautomated DW methods apply electrodes over flex-frameslocal hardware to control & digitize data (leverage existing design)local hardware to control & digitize data (leverage existing design)

• Multi-tiered embedded SHM/NDE systemAE event detected by real-time monitoring of sparse DW electrodesAE event detected by real time monitoring of sparse DW electrodesresistance measurements narrows down affected regionCNT-enhanced thermography used for detailed NDE imaging on ground

6 mm 256 MUX channels50 MHz DAQ


40 mm

Phase I Summary

• CNT-enhanced laminates were fabricated, patterned with silver ink electrode grid, & subsequently subjected to impact damage

in-plane & through-thickness electrical resistance were collectedlittle to no change in values at points away from the damage zone clear changes were observed in both sets of data close to impacted zoneclear changes were observed in both sets of data close to impacted zone high-resolution full-field representation of sub-surface damage presented

D t t th t ti l f i thi h SHM• Demonstrates the potential of using this approach as an SHM solution, with added benefit of CNTs reinforcing the structure

proof-of-concepts for alternative SHM/NDE methods demonstratedp pPhase I results provide an excellent foundation for Phase II research


Phase II Tasks 1 & 2

• Optimizationoptimize method developed in Phase I prior to expanding technologyfind trade between detection resolution & electrode spacinguncover any limitations of this methodalternative approaches to networking & multiplexing to provide bestalternative approaches to networking & multiplexing to provide best detection capability over largest coverage area with least system impact

• Maturation• Maturationmature optimized method by introducing non-ideal factorsreal aerospace structures are exposed to temperature, humidity & loadsexperimental data will be collected for a range of realistic conditionsdevelop compensation algorithms to provide meaningful damage characterization results without false positives in true environmentsp


Phase II Tasks 3 & 4

• Integrationintegrate additional multifunctional capabilities into the systemdevelop unified infrastructure to facilitate resistance measurements, acoustic emission & provide enhanced thermographic NDE imagingsingle multi-physics system will be demonstrated that utilizes CNTs to g p y yreliably & accurately detect, localize & characterize damage in composite

• Expansion• Expansionexpand developments of prior tasks to be demonstrated at a larger scalepractical system will need broader coverage with comparable capabilitydemonstrate technology using a larger panel size (1/2 m), CFRPdemonstrate in representative built-up configuration (integrally stiffened)


CNT-Enhanced Composites for Ice

• Goal to develop a multi-role system for composite rotor bladesdetection of presence of ice on bladesremoval of ice from blades and/or prevention of ice (re)formationdetection/characterization of structural damage on blades

• Current solutions provides high false positive & failure rates

• Novel technologies combine to produce robust/reliable solutioncarbon nanotubes (CNTs)direct write (DW)micro-instrumentationmicro instrumentation

• NAVAIR funded SBIRBAMSBAMS H-53K


Proposed Smart-Blade System

• 3 main system elementsCNTs embedded in composite bladesinterdigitated DW electrodes & thermocouples

Electrically insulating layer

l t dinterdigitated DW electrodes & thermocouplesthin electrically insulating abrasion layer

• System benefitsDirect write electrodes

over electrodes

System benefitsheating elements are structuralsensing elements are structuralno wires requiredno wires requiredconformal (light & low profile)uniform surface coveragevery efficient (conduction only loss) CNT compositevery efficient (conduction only loss)closed-loop feedback possiblecan improve blade impact resistancesolid state (no moving parts)

C co pos te

solid state (no moving parts)


De/Anti-Icing Demonstration

~1 kW/m2

~5 kW/m2


Ice-Detection Formulation

0 4

0.5

0.6

0.7

min bars

20 Points From ‐15C

0 1

0.2

0.3

0.4

Slop

e w/m

ax &

m

0

0.1

No Water 1 mm Water 2 mm Water 3 mm Water

• Effective heat capacity methodconstant current applied to sample for fixed temperature recording timeexponential temperature rise (linear data fit), compared to no-ice case

• Very repeatable resultsshallower slope correlates to more ice on specimenshallower slope correlates to more ice on specimendata consistent for multiple detection temperatures


Acknowledgments

• This research was sponsored by SBIR/STTR fundingAFOSR FA9550-09-C-0165 “SHM of CNT-Enhanced Composites”NAVAIR N68335-10-C-0227 “Composite Self-Monitoring Anti/De-icing” NASA NNX09CC57P “Cable-Free Sensor-Bus for SHM of Composites”

FFRP composites developed by MIT’s NECST Consortium Nano-Engineered Composite aerospace STructures consortiumsupported by large aerospace OEMs & composite fabricatorsAirbus, Boeing, Embraer, Lockheed, Saab, Spirit AeroSystems, Textron, Composite Systems Technology & TohoTenax


fabrication of a multi-physics integral structural ... · ¾direct-write (dw) technology can be...

Documents