Long term stability analysis of a guided wave SHM system for - - PowerPoint PPT Presentation

▶
long term stability analysis of a guided wave shm system
SMART_READER_LITE
LIVE PREVIEW

Long term stability analysis of a guided wave SHM system for - - PowerPoint PPT Presentation

Oct 04, 2022 376 likes •587 views

Long term stability analysis of a guided wave SHM system for platelike structures V. Attarian, F.B. Cegla, P. Cawley Non-Destructive Testing Group Department of Mechanical Engineering Imperial College London SW7 2AZ United Kingdom Outline

slide-1
SLIDE 1

Long term stability analysis of a guided wave SHM system for platelike structures

  • V. Attarian, F.B. Cegla, P. Cawley

Non-Destructive Testing Group Department of Mechanical Engineering Imperial College London SW7 2AZ United Kingdom

slide-2
SLIDE 2

NDE Group 2/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Introduction

– Motivation – Background – Project goals

  • Ruggedized SHM system development
  • Testing
  • Results
  • Conclusions
  • Questions

Outline

slide-3
SLIDE 3

NDE Group 3/19

  • V. Attarian

NDE Group

  • V. Attarian

Introduction

Motivation

slide-4
SLIDE 4

NDE Group 4/19

  • V. Attarian

NDE Group

  • V. Attarian

(1) Suitability for large area inspection (2) Defect sensitivity (3) Complex feature effects

Guided wave transducers in sparse arrays

Background

Features Sensors

slide-5
SLIDE 5

NDE Group 5/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Feature density high in 2D
  • Environmental variability effects

– E.g. Temperature induced delays

Issues for monitoring platelike structures

Background

Baseline subtraction Compensation strategies

100 200 300

TIME (µs) AMPLITUDE TIME DOMAIN SIGNALS

100 200 300

TIME (µs) AMPLITUDE RESIDUAL

NO DEFECT WITH DEFECT

DEFECT DEFECT

slide-6
SLIDE 6

NDE Group 6/19

  • V. Attarian

NDE Group

  • V. Attarian

100 200 300 400

  • 80
  • 60
  • 40
  • 20

RELATIVE TO AMPLITUDE OF 1st ARRIVAL TIME (µs) dB

  • 40 dB

Optimal Baseline

  • Optimal baseline subtraction
  • Baseline signal stretch

Temperature compensation strategies

Background

T oC After stretch

100 200 300 400

TIME (µs) AMPLITUDE

Defect

OPTIMAL BASELINE (AFTER STRETCH) WITH DEFECT RESIDUAL

100 200 300 400

TIME (µs) AMPLITUDE

Defect

100 200 300 400

RESIDUAL TIME (µs) AMPLITUDE

Defect

slide-7
SLIDE 7

NDE Group 7/19

  • V. Attarian

NDE Group

  • V. Attarian

100 200 300 400 500

  • 60
  • 50
  • 40
  • 30
  • 20

dB TIME (µs) 1st ARRIVAL DEFECT

Optimal Baseline

Defect detection in container panel

Background

X (M) Y (M) 0.5 1 0.5 1 1.5 2

  • 50
  • 45
  • 40
  • 35
  • 30

dB

X Y

Clarke et al (IEEE 2009)

T oC Compensated baseline subtraction x x

1 4

Triangulating with all sensor pairs

3

Ellipse imaging

2

slide-8
SLIDE 8

NDE Group 8/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Long term testing of structure in exposed weather

– Need to ruggedize transducer

  • Assess feasibility and variability of reliable monitoring

– Robustness – Baseline stability

  • Study on simple plate and complex structure

Performance in real life scenario?

Project goals

slide-9
SLIDE 9

NDE Group 9/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Phase stability

Finite element analysis

  • Pure mode excitation
  • Strain relief
  • Waterproof

Cap

  • Electrical connectivity/

grounding

Transducer design

Ruggedized SHM system development

Epoxy

+V

slide-10
SLIDE 10

NDE Group 10/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Want to excite PZT

– Strong S0 – Off PZT resonance (outside red region) – High S0/A0 ratio (green)

50 100 150 200

  • 10
  • 5

5 10 15 20 8 dB

Ruggedized SHM system development

Not to scale

D/2 vA0 vS0 vPZT d = 300 mm

50 100 150 200 20 40 60 80 100 kHz x10-6 S0 A0

Excitation Frequency (kHz)

Energy in S0 to A0(dB) Velocity FRFs (m/s/V)

5 mm

slide-11
SLIDE 11

NDE Group 11/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Undesirable coupling of portal frame dynamics in excited S0
  • Parametric + material changes allow shift to and damping

Ruggedized SHM system development

h

100 200 300

  • 0.05

0.05 ( ) UNENCAPSULATED 1ST PROTOTYPE CAP

Minimize cap reverberations

In-plane Velocity (mm/s) Time (µs)

2 mm

100 200 300

  • 0.05

0.05 UNENCAPSULATED FINAL CAP

Time (µs) In-plane Velocity (mm/s)

slide-12
SLIDE 12

NDE Group 12/19

  • V. Attarian

NDE Group

  • V. Attarian

Final transducer

Ruggedized SHM system development

slide-13
SLIDE 13

NDE Group 13/19

  • V. Attarian

NDE Group

  • V. Attarian
  • 10

10 20 30 40 50

  • 10

10 20 30 40 50

  • 10

10 20 30 40 50

  • 10

10 20 30 40 50

  • Baselines collected after 3 months of outdoor robustness testing
  • Signals acquired at 0.5oC increments in [-5, 40]oC
  • Noise floor imaged with 0.5oC gap to nearest baseline throughout range

0.5 1 0.2 0.4 0.6 0.8 1 1.2

  • 50
  • 48
  • 46
  • 44
  • 42
  • 40
  • 10

10 20 30 40 50

  • 10

10 20 30 40 50

Baseline database collection

Testing

0.5 1 0.2 0.4 0.6 0.8 1 1.2

  • 50
  • 48
  • 46
  • 44
  • 42
  • 40

d

<-40 dB  Sufficiently low noise for SHM

X (M) Y (M) dB dB X (M) Y (M) SETPOINT (oC) PLATE TEMPERATURE (oC)

slide-14
SLIDE 14

NDE Group 14/19

  • V. Attarian

NDE Group

  • V. Attarian

Instrumented plate outdoors

Testing

transducers

thermocouples

Multiplexer

Tx Rx

High pass filter

A/D

Waveform generator PC control

  • Monitoring since beginning of March for +4 months
slide-15
SLIDE 15

NDE Group 15/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Tx-Rx ok for yr through temperature swings (-10 to 30oC), rain,

light snow

  • Signal amplitudes

– >80% pairs show △amplitude

  • To -37% initial value, on average
  • Unrelated to temperature

– Regressions on noise levels/pair indicate correlation with △amplitude

Robustness

Results

20 40 60 80 100

  • 40
  • 30
  • 20
  • 10

SENSOR PAIR TRANSMITTER 1 - RECEIVER 2 DAYS ∆ AMPLITUDE RELATIVE INITIAL (%) DATA FIT

slide-16
SLIDE 16

NDE Group 16/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Imaging suggests issues for monitoring small reflectors

Baseline stability assessments

+1 week +3 weeks +6 weeks +11 weeks +14 weeks

Results

0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2 g ; ; 0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2

0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2

X (M) Y (M) Statistics from within area

  • 30
  • 40
  • 35

0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2

dB

  • 30
  • 40
  • 35 dB

Noise level in area

20 40 60 80 100

  • 50
  • 45
  • 40
  • 35
  • 30
  • 25

TIME (DAYS)

slide-17
SLIDE 17

NDE Group 17/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Statistical metric suggests noise exhibits lack of sensitivity to

location in coverage area

Spatial distribution of noise

Results

0.2 0.4 0.6 0.8 1 1.2 0.2 0.4 0.6 0.8 1 1.2

ri ro

X (M) Y (M) dB

  • 30
  • 40
  • 35

dB

0.2 0.4 0.6

  • 50
  • 45
  • 40
  • 35
  • 30
  • 25

R (M) M (dB) WEEK 1-2 WEEK 12 WEEK 8

slide-18
SLIDE 18

NDE Group 18/19

  • V. Attarian

NDE Group

  • V. Attarian
  • Key findings

– Rugged transduction of GW

  • feasible for >6 months
  • amplitudes decline over time identified

– Noise floor changes quantified

  • drifts to -25 dB in +4 months observed + analyzed
  • trends don’t exhibit spatial dependency
  • Future work

– Continued monitoring outdoors/analysis – Fix amplitude decline issue – Quantifying defect detection sensitivity over time – Further prototyping of array on complex structure

Conclusions

Possibility of frequent data collection and detrending

  • A. Galvagni talk

@ 11:30 AM

slide-19
SLIDE 19

Transducer @ ½ yr

Thank you. Questions?