Embedded Sensing: Piezoelectric Active Sensing & Guided Waves

Transcription

1 Structural Health Monitoring Using Statistical Pattern Recognition Embedded Sensing: Piezoelectric Active Sensing & Guided Waves Presented by Eric B Flynn, Ph.D. Outline Modeling Signal processing for Guided Waves Nonlinear Acoustics/Time Reversal Acoustics Integration with other SHM technologies Commercially Available Systems Applications Full Wavefield Imaging Research Issues #6. Guided waves for SHM: 2

2 Many structures in SHM behave as waveguides, i.e. a plate Excite longitudinal plane wave Traction free boundary Reflect a longitudinal plane wave and a shear plane wave Lamb waves are a superposition of all of the longitudinal and shear waves propagating down a plate. #6. Guided waves for SHM: 3 Multiple modes exist in Lamb wave propagations 1 st symmetrical mode (S 0 ) 1 st antisymmetrical mode (A 0 ) cg (m/s) At any frequency, if we excite ONE pulse (wave), there will be MULTIPLE pulses (at least two) traveling through the wave guide freq (khz) #6. Guided waves for SHM: 4

3 Lamb waves are also dispersive Different frequencies travels at different speeds. Non-dispersive Dispersive Figures from #6. Guided waves for SHM: 5 Multi-mode and dispersion Waves spread out in time as they travel Imposes difficulties in signal processing in some cases Selection of optimal frequency is needed to reduce these effects Normalized amplitude Time (μs) Excitation Multimodes Normalized amplitude Time (μs) Received signal dispersion #6. Guided waves for SHM: 6

4 Multi-mode and dispersion Transducer #6. Guided waves for SHM: 7 How to select the excitation frequency? 0 1 Mhz-mm must be used to excite only A 0 and S 0 modes. Below 800 khz or so, A 0 mode will be reasonably non-dispersive Notice that A 0 modes travels much more slowly than S 0 mode -> wave length (V/f) is smaller than So A 0 will be more sensitive to small defects in a structure while So is ideal for reflection/ scattering estimation. (frequency thickness product) #6. Guided waves for SHM: 8

5 Modeling of wave propagations Theory behind Lamb wave propagations on homogenous, isotropic materials in simple geometry is relatively well understood Cawley s group (Imperial college), Rose (Penn State), Giurgiutiu (South Carolina), Cesnik (Michigan), Lanza (UCSD), Staszewski (U of Sheffield), Mall (UCLA), to name a few Ranges from commercial FEM codes (ABQUS etc), analytical estimation, Semi Analytical FE, or customized software (Disperse) Normalized Amplitude Model 1 Exp Time(ms) Puckett (2008) for thin AL plate These model have been rarely used in SHM Computationally intensive Difficulties in complex geometry (joint, different BC conditions) Used only as guidelines for deploying sensing systems #6. Guided waves for SHM: 9 The analytical model considers excitation, propagation, and reception. x Excitation Ragahavan and Cesnik, SPIE 2004 u( x, z N S D S S A 0a i t S NS ( k ) (2) S A N A( k ) (2) A b) i e J1 k a H ( k x) J k a H ( k x) S 1 1 A 1 kq( k ( k 2 2 q k S D ( k ) 2 q )cos( pb)cos( qb) 2 ) 2 S cos( pb)sin( qb) 4k 2 k A D ( k ) pasin( pb)cos( qb) A Propagation e ikx Reception A recpt x a 2 rr x a 2 cos( ) rd dr #6. Guided waves for SHM: 10

6 The predicted signals from the analytical model agreed with experiments (with reflections, Puckett 2008). #6. Guided waves for SHM: 11 The predicted signals from the analytical model agreed with experiments (with reflections, Puckett 2008). 1 Normalized Amplitude Model Exp Time(ms) #6. Guided waves for SHM: 12

7 The analytical signal can be decomposed into each reflected path (Puckett 2008) #6. Guided waves for SHM: 13 Transfer functions can be calculated that illustrate over what frequencies the modes are excited (Giurgiutiu et al., Cesnick et al) Normalized Normalized Normalized Magnitude Magnitude Magnitude Total Transfer Function for 1/4" PZT Reception Transfer Function for 1/4" PZT Excitation Transfer Function for 1/4" PZT S0 mode S0 mode A0 mode A0 S0 mode A0 mode Frequency (khz) #6. Guided waves for SHM: 14 Los Alamos DynamicsFrequency Structural Dynamics (khz) and Mechanical Vibration Consultants Courtesy of A. Puckett (2008)

8 Signal Processing for Guided Waves On of hottest research topics to improve damage detection though advanced signal processing. Many techniques are out there, but in general they can be categorized by the feature they use Reflection Deflection/scatter Attenuation/Shape distortion Each Feature shows different sensitivities to different types of damage #6. Guided waves for SHM: 15 The first step usually involves Data cleansing Improving SNR is important Redundant averaging Wavelet analysis is increasingly used for this purpose Courtesy of Rizzo and Lanza di Scalea (2004) #6. Guided waves for SHM: 16

9 Reflections Damage will cause a reflection at the site, and this reflected wave will be captured at the actuation site. Giurgiutiu et al Location of damage can be identified with triangulation using an array of sensors. Several processing algorithms available to estimate TOF Works well with isotropic materials Reflected wave must be at least 10% of boundary reflected wave to be reliably detectable. #6. Guided waves for SHM: 17 Attenuation/Shape distortion (Sohn et al. 2004) 7 7 Impact location Impact 33.0 m/s Selected damaged paths Estimated Delamination area Response time signals corresponding to a damaged path (from PZT 7 to PZT 14) Voltage (mv) Time Points #6. Guided waves for SHM: 18

10 Attenuation/Shape distortion Wavelet coefficients or frequency spectrum are typically used as an damage sensitive features Chang (Stanford) and many others Pros Signal processing is relatively straightforward Works well with non-isotropic materials Less affected by boundary reflected waves Ideal for monitoring critical locations (hot spots) Cons Dense arrays of sensors are usually required Sensor locations are critical. Difficult to cover boundary regions #6. Guided waves for SHM: 19 Scattering Scattering features may be used to create image caused by damage (Wilcox (Bristol), Flynn (UCSD), Chang (Stanford) ) Scattered signals = Data w/damage Baseline There are many different ways of doing this (Hilbert transform, multiple baselines etc) Line of Possible damage location Todd and Flynn (2008) #6. Guided waves for SHM: 20

11 Scattering Pros Can visualize damage with the limited number of sensors Sensor can cover relatively large areas. In general, more sensitive than using the attenuation feature (for metallic structures) Cons Other reflected wave (boundary etc) causes problems May not work well if the plates are not isotropic Many sensors are required for sufficient resolution to pinpoint the locations Advanced signal processing techniques are required to enhance SNR #6. Guided waves for SHM: 21 Phase Array directional tuning An array of transducers used to provide directionality of propagated waves Giurgiutiu (South Carolina), Adams (Purdue), Pines (Maryland) and many others Excite and sense on each transducer with a predefined phase delay in order to steer the wave in a particular direction. Time delay Pros Improved signal SNR Allows strategic deployment of sensors More precise localization Cons Works well in isotropic materials with regular cross-section Depends on wave reflections or scattering Less coverage for same transducer count #6. Guided waves for SHM: 22

12 Phase Array example Figures courtesy of Giurgiutiu, 2002 #6. Guided waves for SHM: 23 Circular Array Node Phase Array example 1 Actuator 6 Sensors Multi-site damage detection 1 cm disc magnets 91 cm Figures courtesy of Metis Design, 2011 #6. Guided waves for SHM: 24

13 On Signal Processing of guided waves Many advanced signal processing techniques were adopted for data cleansing, data normalization, and enhanced damage detection time frequency analysis, wavelets, Hilbert transformation, filters etc. Pattern recognition methods using neural networks, support vector machines were also explored - Worden (U. of Sheffield) Nonlinearity detection using modulation, time reversal acoustics were also explored. With and without Fatigue crack, Staszewski et al (2009) #6. Guided waves for SHM: 25 Nonlinear Acoustics Harmonic Distortion Method: measure the degree of the nonlinear (harmonic) distortion of a sinusoidal acoustic signal Usually use 2 nd harmonic as the magnitude of higher harmonics rapidly drops off Used for detection of fatigue cracks, adhesive joints, cracks in concrete Sensitive to the small defects, however must be able to distinguish equipment/electronics (harmonic) nonlinearity Figure Courtesy of Donskoy (2008) #6. Guided waves for SHM: 26

14 Nonlinear Acoustics Modulation Method: utilize the effect of the nonlinear interaction of acoustic/vibration waves in the presence of the nonlinear defects. Vibration modulation Impact modulation Not affected by equipment nonlinearity, enhanced sensitivity In order to achieve repeatability and high sensitivity, the modulating stress applied to the damaged area should be constant Figure Courtesy of Donskoy (2008) #6. Guided waves for SHM: 27 Time Reversal Acoustics A source emits waves that travel through a medium, and detected by one or more receivers. The detected signals are then reversed-in-time and rebroadcast from their respective receiver positions (Time Reversal Mirror) The back-propagating waves simultaneously arrive at the original source location in phase, producing a time reversed focus, which is a reconstruction of the original source Figure Courtesy of Anderson et al. (2008) #6. Guided waves for SHM: 28

15 Time Reversal Acoustics TRA relies on the principle of spatial reciprocity, which is not affected by dispersion, scattering, anisotropy, refraction or linear attenuation. The spatial reciprocity is broken when the nonlinear elastic effect is present - Amplitude dependent attenuation - Nonlinear cracks - (Significant) Delamination in composite plates #6. Guided waves for SHM: 29 Time Reversal Mirror TRMs are made of large transducer arrays, allowing the wave field reflected by a defect to be sampled, time reversed and re-emitted. TRM allows one to convert a divergent reflected wave issued from a defect into a convergent wave focusing on this defect. Improves the focusing of the ultrasound, particularly for samples with complex shapes Makes it possible to detect small defects with a greater degree of certainly #6. Guided waves for SHM: 30

16 Crack imaging with TRA #6. Guided waves for SHM: 31 Integration with other SHM approaches With impedance methods: Kabeya et al. (1998), Monnier et al. (2000), Giurgiutiu et al, Wait et al, (2004) Thien et al. (2007) With Acoustic Emission techniques: Blanas and Das-Gupta (1999) Accellent, Giurgiutiu et al., and many others Todd s group at UCSD: chaotic ultrasonic excitation for guided waves #6. Guided waves for SHM: 32

17 Integrated pipe monitoring system (Thien et al. 2005) MFC Connection damage in joints: Impedance methods Cracks, corrosion in pipeline: Lamb wave propagation Both methods utilize the same MFC patches #6. Guided waves for SHM: 33 Impedance method for joint monitoring Damaged Joint 1-5: Baselines 6,7: Loosened 1 bolt #6. Guided waves for SHM: 34

18 Lamb wave propagation for crack/corrosion detection Pulse-Echo Wave generation Boundary reflection Voltage Time x 10 4 #6. Guided waves for SHM: 35 Both attenuation and reflection features are used to detect and locate surface damage Pulse-Echo Damage Wave generation Reflection from damage reflection Boundary reflection base dam 1 attenuation Voltage Tim e x 10 4 Time of flight #6. Guided waves for SHM: 36

19 Commercial Systems Accellent Company Smart Layer Technology Rack DAQ Hardware Metis Design Corp. Phased-Array Technology Integrated DAQ Hardware #6. Guided waves for SHM: 37 Application of Guided Wave SHM The most successful application of guided waves in SHM have been pipelines (Cawley s group, Rose s group) Made the transition to commercialized products. coefficient amplitude 1 baseline 0.5 damage 0 (a) time [ms] absolute difference signal difference threshold limit time [ms] (b) Figures courtesy of Guided Ultrasonics Ltd time [ms] percent difference [%] signal difference threshold limit #6. Guided waves for SHM: 38 (c)

20 Aerospace Applications (Chang, Accellent) Fuselage Lap Joints Bonded Repair Patch Composite repair layers Actuator Strip Rivets sensor network Adhesive film Rivets SMART layer SMART Patch Assembly Sensor Strip Crack #6. Guided waves for SHM: 39 Aerospace Applications Many studies are being performed in the application of guides waves to aerospace structures Studies also include The effects of adhesive materials (Accellent, AFRL) High temperature applications (Cesnik, Giurgiutiu) Long term reliability (Giurgiutiu, Cesnik, AFRL) Space applications (NASA, LANL) Dalton et al (2001): Guided wave based methods are ideal for the free and tapered skin configurations and lap joints (simper configuration). However, it would be challenging for complicated configurations or skin coated with sealant (due to damping) #6. Guided waves for SHM: 40

21 Naval Applications (Metis Design) Crack 183 cm Loose Bolt Digital Com Line (Kessler & Flynn 2011) #6. Guided waves for SHM: 41 Civil Structures Lanza di Scalea (2004): Monitoring pre-stress and cracks in multi-wire steel strands using guided waves, monitoring of defects in rail Park S. et al (2006): Monitoring of bridge girder integrity Many others Figures Courtesy of Lanza di Scalea UCSD Figure Courtesy of Park, S. et al. (2006) #6. Guided waves for SHM: 42

22 Full Wavefield Imaging (Hybrid SHM/NDE) Transducer Senses Transducer Excites QSL Excites LDV Senses Excite with pulsed-laser Pulsed excitation Sense with embedded transducer Measure transient response Excite with fixed transducer Continuous harmonic excitation Sense with laser Doppler vibrometer Measure steady-state response #6. Guided waves for SHM: 43 Pulsed Excitation QSL (Excites) Scan time: 15 minutes Raw Local Dispersion Eric Flynn, See-Yenn Chong & Jung-Ryul Lee #6. Guided waves for SHM: 44

23 Steady Excitation Transducer Excites LDV Senses Scan time: 8 seconds Raw Eric Flynn & Gregory Jarmer Wavenumber m Local Wavenumber #6. Guided waves for SHM: 45 Research Issues Although guided waves have shown great potential, transitioning to real-world applications is still challenging Many research issues remain, including Establishment of design guidelines Advanced modeling (PZT-structure interaction, more complex geometries) Sensor location optimization (must be tied with specific signal processing techniques) User-friendly software Verification of performance against traditional NDE methods Compensation of environmental variability Long term reliabilities of sensors Do not oversell the performance of the guided waves #6. Guided waves for SHM: 46

24 Useful References Thien, A.B., Chiamori, H.C., Ching, J.T., Wait, J.R., Park, G. 2008, The use of Macro-Fiber Composites for Pipeline Structural Health Assessment, Structural Control and Health Monitoring, Vol. 15, pp Raghavan, A., Cesnik, C.E.S., 2007, Review of Guided-wave Structural Health Monitoring, The Shock and Vibration Digest, Vol. 39, pp Rose, J. L. (1999). Ultrasonic Waves in Solid Media, Cambridge University Press, Cambridge Wilcox P.D., Lowe M.J.S. and Cawley P., Mode and transducer selection for long range Lamb wave inspection, Journal of Intelligent Material Systems and Structures, v. 12, p , August 2001 Diaz Valdes S.H. and Soutis, C., Health monitoring of composites using Lamb waves generated by piezoelectric devices, Plastics, Rubber and Composites, v. 29 (9), p , 2000 Kessler S. and Spearing M., Design of a piezoelectric-based structural health monitoring system for damage detection in composite materials, Proceedings of the SPIE, v. 4701, p , 2002 Raghavan A. and Cesnik C.E.S., Modeling of piezoelectric-based Lamb-wave generation and sensing for structural health monitoring, SPIE Symposium on Smart Structures & Materials/ NDE 2004, Paper , San Diego, California, Mar , 2004 Giurgiutiu V., "Lamb wave generation with piezoelectric wafer active sensors for structural health monitoring," Proceedings of the SPIE, v. 5056, p , 2003 Rizzo P. and di Scalea F.L., Discrete wavelet transform to improve guided wave-based health monitoring of tendons and cables, Proceedings of the SPIE, v. 5391, p , 2004 Ihn J.-B. and Chang F.-K., Detection and monitoring of hidden fatigue crack growth using a built-in piezoelectric sensor/actuator network: I. Diagnostics, Smart Materials and Structures, v. 13, , 2004 Sohn H., Park G., Wait J.R., Limback N.P., and Farrar C.R., Wavelet based active sensing for delamination detection in composite structures, Smart Materials and Structures, v. 13, p , 2004 Ing R.K. and Fink M., Time recompression of dispersive Lamb waves using a time reversal mirror Application to flaw detection in thin plates, Proceedings of the IEEE Ultrasonics Symposium, v.1, p , 1996 #6. Guided waves for SHM: 47 Useful References Kessler S.S., Spearing M.S., Shi Y., and Dunn C., Packaging of structural health monitoring components, Proceedings of the SPIE, v. 5391, p , 2004 Dalton, R. P., Cawley, P., and Lowe, M. S. J. (2001). The Potential of Guided Waves for Monitoring Large Areas of Metallic Aircraft Fuselage Structure, Journal of Nondestructive Evaluation, 20(1): Purekar A. and Pines D., Damage interrogation using a phased piezoelectric sensor/actuator array: simulation results on two dimensional isotropic structures, Proceedings of the 44 th AIAA/ASME/ASCE/AHS Structures, Structural Dynamics, and Materials Conference, Norfolk, Virginia, Paper , 7-10 Apr Sundararaman S. and Adams D.E., Phased transducer arrays for structural diagnostics through beamforming, Proceedings of the American Society Lin M., Qing X., Kumar A., and Beard S.J., SMART layer and SMART suitcase for structural health monitoring applications, Proceedings of the SPIE, v. 4332, p , 2001 Yang J., Chang F.-K., and Derriso M.M., Design of a built-in health monitoring system for thermal protection panels, Proceedings of the SPIE, v. 5046, p , 2003 Lanza di Scalea, F., Matt, H., Bartoli, I., Coccia, S., Park, G., Farrar, C.R., 2007, Health Monitoring of UAV Wing Skin-to-Spar Joints using Guided Waves and Macro-Fiber Composite Transducers, Journal of Intelligent Material Systems and Structures, Vol. 18, No.4, pp Giurgiutiu V., Zagrai A. and Bao J., Damage identification in aging aircraft structures with piezoelectric wafer active sensors, Journal of Intelligent Material Systems and Structures, v. 15, p , 2004 Lee, B. C., and Staszewski, W. J. (2003). Modelling of Lamb Waves for Damage Detection in Metallic Structures: Parts I and II, Smart Materials and Structures, 12(5): Lee, B. C., Manson, B., and Staszewski, W. J. (2003). Environmental Effects on Lamb Wave Responses From Piezoceramic Sensors, Materials Science Forum, : Anderson, B. E., M. Griffa, C. Larmat, TJ Ulrich, and P. A. Johnson, Time Reversal, Acoust. Today (2008). D. M. Donskoy Nonlinear Acoustics Methods, Encyclopedia of Structural Health Monitoring, #6. Guided waves for SHM: 48

25 Useful References Park, S., Yun, C.B., Roh, Y., Lee, J.J., 2006, PZT-based active damage detection techniques for steel bridge components, Smart Materials and structures, 15, Monnier, T., Jayet, Y., Guy, P., and Baboux, J. C. (2000). Aging and Damage Assessment of Composite Structures Using Embedded Piezoelectric Sensors, in Review of Progress in Quantitative NDE, Vol. 19, D. O. Thompson and D. E. Chimenti, eds, Plenum Press, New York, Blanas, P., and Das-Gupta, D. K. (1999). Composite Piezoelectric Sensors for Smart Composite Structures, in Proceedings of the 10 th International IEEE Symposium on Electrets, European Cultural Centre of Delphi, Greece, Flynn EB, Todd MD, Wilcox PD, Drinkwater BW, Croxford AJ. Maximum-likelihood estimation of damage location in guided-wave structural health monitoring. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Science. 2011;467(2133): Croxford A., Wilcox P., Drinkwater B., Konstantinidis G. Strategies for guided-wave structural health monitoring. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Science Nov 8;463(2087): Flynn EB, Kessler SS, Todd M. Identifying Scatter Targets in 2D Space using In Situ Phased Arrays for Guided Wave Structural Health Monitoring. Proceedings of the 8th International workshop on Structural Health Monitoring. Palo Alto, CA; Yinghui Lu, Michaels JE. Feature Extraction and Sensor Fusion for Ultrasonic Structural Health Monitoring Under Changing Environmental Conditions. Sensors Journal, IEEE DOI /JSEN ;9(11): E. B. Flynn, S. Y. Chong, G. J. Jarmer, and J. R. Lee, Imaging structures through local wavenumber estimation of guided waves, NDT&E International, #6. Guided waves for SHM: 49