Recent advances in thermal wave detection and ranging for non-destructive testing and evaluation of materials

Ravibabu Mulaveesalaa,c*, V.S.Ghalib,c, Vanita aroraa, Juned.A.Siddiquic, Amarnath.Muniyappac and

Masahiro Takeid a Department of Electrical Engineering., Indian Institute of Technology Ropar, India;

b Signal Processing Research Group, K L University, Green Fields, Vaddeswaram, Guntur (Dist.), Andhra Pradesh, 522 502, India.;

c InfraRed Imaging Laboratory (IRIL), Electronics and Communication Engineering Research Group, PDPM-Indian Institute of Information Technology Design and Manufacturing, Jabalpur,

Airport road, Khamaria (P.O), Jabalpur, India-482005; d Graduate School of Chiba University, Artificial System Science, 1-33 Yayoi Inage Chiba #263-

8522 Japan.

ABSTRACT Thermal Wave Detection and Ranging (TWDAR) for non-destructive testing (TNDT) is a whole field, non-contact and non-destructive inspection method to reveal the surface or subsurface anomalies in the test sample, by recording the temperature distribution over it, for a given incident thermal excitation. Present work proposes recent trends in non-stationary thermal imaging methods which can be performed with less peak power heat sources than the widely used conventional pulsed thermographic methods (PT & PPT) and in very less time compared to sinusoidal modulated Lock-in Thermography (LT). Furthermore, results obtained with various non-stationary thermal imaging techniques are compared with the phase based conventional thermographic techniques. Keywords: Active thermography, Pulse compression, Frequency modulated thermal wave imaging, Barker code.

1. INTRODUCTION Thermal Non-destructive Testing (TNDT) has become a favorite choice among the various widely used subsurface imaging methods due to its whole field, non-contact and nondestructive evaluation procedure to assess the structural integrity of the objects. It has been carried using either in passive or active approaches. In passive approach inherent thermal response of the test object is captured and analyzed to reveal subsurface features. Capturing the thermal response in the absence of any external stimulation limits its applicability in producing better contrast for defects located at deeper depths in the test specimen. Whereas, in active approach, with its controlled heat stimulus along with its well supported processing techniques makes it as a reliable qualitative and quantitative testing procedure for surface and subsurface defect detection. In this method, a controlled stimulus is given and temperature map of the test object surface is captured using infrared imager. Thermal waves generated from the constituent frequency components of the stimulation in addition to various processing methods facilitate better depth analysis and fine subsurface details.

Among various excitation methods, Pulsed Thermography (PT), Lock-in Thermography (LT) and Pulse Phased Thermography (PPT) are widely used in numerous thermographic applications. In PT1, a short pulsed, high peak power stimulus is given to the sample and temporal temperature map has been captured. Influence of non-uniform emissivity and non-uniform heating over the surface may result in erroneous predictions in this direct contrast method and limits its applicability in spite of its quickest evaluation. In modulated wave thermography i.e., LT2and Frequency modulated thermal wave imaging3,4 (FMTWI) make use of low peak power sources for longer durations and improve penetration of thermal waves. Less attenuation provided by low frequency thermal waves has been used by LT to provide deeper depth details. But mono frequency excitation may not better resolve the defects at different depths and demands repetitive experimentation using a number of frequencies in realistic applications. In PPT5, test object is stimulated as it was done in PT, but analysis is carried by the application of Fast Fourier Transform (FFT) over thermal profiles of each pixel and detection can be carried from phasegrams at different frequencies.

Thermosense: Thermal Infrared Applications XXXV, edited by Gregory R. Stockton, Fred P. Colbert, Proc. of SPIE Vol. 8705, 870510 · © 2013 SPIE · CCC code: 0277-786X/13/$18 · doi: 10.1117/12.2018465

Proc. of SPIE Vol. 8705 870510-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/23/2013 Terms of Use: http://spiedl.org/terms

a

11'

FMTWI makproperties and

This non staticontributes folocation for tspecimen can

Where T(x,t) and ‘α’ is the The solution location of de

Where =σ

These thermausing phase b(PC). Applicadelay betweenIn order to prdata of the costationary exprocessed usi

Figure 1.a After repeatincontrast in phdetectability dimensions afrequency swanalysis using

2. FRkes use of stimd its dimensio

ionary stimulaor temperaturthe given stim

n be obtained b

is the temper thermal diffuof these non

epth ‘x’ as

sα

.

al waves produbased analysiability of pulsn their responrocess the caporresponding pxcitation can bing either FT o

a.. Thermogra

ng the abovehasegrams or cin the above

as shown in Fiweep 0.01Hz tog FT( Figure.2

EQUENCYmulation with ons is as given

y(t) Sin [2=ated excitation

re variations omulation withby solving the

rature responsusivity of the on-stationary w

uces a correspis by the applse compressionses similar to ptured thermapixel as shownbe removed uor PC as show

am sequence (

e process on tcorrelation im

e processes, eigure.2.a witho 0.1Hz. It is2.c) for FMTW

Y MODULcontinuous m

n by 02 (f b t)t];π +

n initiates simover the objech a homogenoe heat equation

e of the objecobject.

waves is obtai

T(x,s)k

=σ

ponding radiatlication of Fo

on approach foRADAR naml history, temn in Figure 1.using a suitab

wn in Figure 1

n thermogram

temperature pmages can be uexperimentatioh air backing (s observed thaWI.

LATED THmodulated freq

0 f =milar thermal wct surface at tous solid in thn in transform

2

2T 1=

α x

∂ ∂∂∂

ct at a depth ‘x

ined in transf

Q(s)2 L(1 e )− σ−

tion which caourier transforor thermal wa

med thermal wmperature prof

a, temporal thble fitting pro.b.

ms) Figure 1

profiles of allused for idention is carried(b,c,d) and copat PC (Figure.

HERMAL Wquency sweep

Initial frequenwaves which the defect loche absence of

m domain unde T t∂

x’ from the ob

form domain,

x (x[e e−σ σ+

an be capturedrm(FT) or usiaves facilitateswave detectionfiles of each phermal offset docedure. Thes

.b. Processed

l the pixels inification of suon a CFRP pper coated b.2.b) processin

WAVE IMAchosen accor

ncy; b= Sweefurther propagations. The tef any heat souer suitable bou

bject surface a

for a sample

T2 L) 0]s

− −

d by infrared cing correlations identification

n and ranging 6pixel were condue to the mease offset rem

correlation se

n the field of ubsurface anomsample contaacking (a,e) ung is perform

AGING rding to the sa

ep rate; gate into the temperature reurce or sink iundary condit

at a given inst

e of thickness

camera and prn based pulse

on of defects o6,7,8 (TWDAR

nstructed froman increased d

moved profiles

equence (2 n -

f view, localizmalies. In ordaining flat bousing chirped

ming better tha

ample therma

(1test object andesponse at anyinside the testions,

(2

tant of time ‘t

s ‘L,’at defec

(3

rocessed eithee compressionon the basis oR). m the temporaduring the nons were further

1 images)

zed parameteer to study thettom holes ostimulation o

an phase based

al

) d y st

)

’

ct

)

r n f

al n r

r e f f d

Proc. of SPIE Vol. 8705 870510-2

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/23/2013 Terms of Use: http://spiedl.org/terms

15.1

-

10.1

a ( )L 008.

t_ 02.2

01.8

re® 02

01.2

8 -I

o.

0.22

0.2!

n '14

0.2!

0.2

-

-0.1 Y>

oIT)

-0.105 8..._O,as

_n 11 T,,

i- a

óU

-0.115 mN

E-0.12 óZ

OO

OW

W-

Nor

mal

ized

cor

rela

tion

coef

ficie

nt

0.403

a)

0.2 2

Figure 2.a.

3. Digitized freqin FMTWI. Icontinuous siDFMTWI is But data acqu Similar to FM

Where Pn =

Detection capExperimentatexperiment oFigure.3.a.

Figure 3.

a Layout of CF

DIGITIZEquency modulIt can be geneine chirp in agreater than i

uisition and pr

MTWI on solv

T(x,

( )n2 Q 10(2n 1)

−

π +,

pability of Dtion makes uson a CFRP sp

a a. Layout of C

RP sample.

ED FREQUlated thermal erated by conaddition to theits continuousrocessing are s

ving heat equa

(2n,s)k

∞

=−∞=σ

∑

thQ is n han

FMTWI can se of DFM stpecimen with

CFRP sample.

Figure 2.b. C

UENCY Mwave imaging

nverting the sieir harmonics wave countesimilar to FM

ation, thermal r

)Pn Q (n22n 1

2 L (1 e )

+− σσ −

armonic coeffi

be analyzed timulation ob

h a set of bac

. Figure 3.b

b Correlation im

MODULATEg 4,9(DFMTWine chirp intos. Thus the baerpart and faci

MTWI.

response for a

s )2n+1

x[e)

−σ

ficient in the o

using phase btained from dck drilled ho

b b. Correlation

mage at 17s

ED THERMWI) is a digitalo binary squarand width andilitate generat

a chosen depth

(x 2 L)eσ −+

orthogonal ba

based and PCdigitization oles of differe

n image at 4.2

Figure 2.c.

MAL WAVcounter part

re chirp. It cod the incidention of stimula

h ‘x’ is analyt

T0]s

−

sis expansion

C based analyf continuous

ent sizes at v

s Figure 3

c Phase image

VE IMAGIof the chirped

onsist all the ft energy are eation from hal

tically express

n of DFM.

ysis as shownchirp used in

various depths

c 3.c. Phase ima

at 0.015 Hz.

ING d version usedfrequencies oenhanced withlogen sources

sed as

(4

n in Figure.3n the previouss as shown in

age at 0.02 Hz

d f h s.

)

. s n

z.

Proc. of SPIE Vol. 8705 870510-3

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/23/2013 Terms of Use: http://spiedl.org/terms

2

1.95 vr1.9

1.85

cc

Ph

90

8070

60

da-

Ix 40cn 30

20

IV

0f

Irrelation Ima e 73

ase !ma ,e 68

1.4

1.2

b c

71 64.2

66.7 58.6

orrelation Image

d e f

51 64.6 62

43.8 53.2 52.1

Defect

Phase Image

h

Modulated expenetration anand their detecompression section highlmodified 7 bmethodology On solving he

T(x, t)

Application oresults in corrSignal to noiscase of defect

Figure 4.a. Cor

In order to cois carried ovepower 1 kW whereas seveschemes for tthe respective

xcitations impnd depth resoection capabilof these schemlights a binarbit Barker cody and compare

eat equation, t

( )( )

34 α βP

)K π x

=

of pulse comprrelation imagse ratio (SNRts ‘d’ and ‘h’.

rrelation image

5. COompare the reser the CFRP each for dura

en bit Bi-phasethe same CFRe scheme as sh

4. PHASEpose more eneolution. FMTWlity is tested umes influencery phase modde. Defect detd it to conven

thermal respon

( )niP 40

12x i 1

−∑=

pression and pe at 8.5s (Figu

R) of the defec

at 8.5s Figur

OMPARAsponse12,13,14 ospecimen of ation of 100se coding is us

RP specimen uhown in Figur

E CODED ergy using lowWI and DFMTusing PC basees their detectidulated thermtection capabi

ntional phase b

nse for this sti

( )3 2t a τ i−

phase based anure 4.a) and p

cts that PC bas

re 4.b. Phase im

TIVE ANAof FMTWI, DFigure 3.a. O. Both FMTW

sed with Barkeunder similar ere 5.

THERMAw peak powerTWI have beeed analysis. Bion sensitivity

mal wave imaility of this scbased imaging

imulation is g

(2-x

4α t aie

−

nalysis over tphase image ased detection

mage at 0.014H

ALYSIS OFDFMTWI and Optical excitatWI and DFMTer coded stimexperimental

AL WAVE Irs with suitablen already dis

But energy shay. In order to raging system1

cheme has beg.

given by

)τiwhere ni

the CFRP samat frequency 0outperformed

z.

F PULSE CBarker coded

tion is given fTWI employ

mulation. Best conditions illu

IMAGINGle bandwidth, cussed in the

aring by compreduce the stre0,11 based on

een analyzed

0,1,2,3; ai=

mple of Figure0.014Hz (Figud the phase ba

Figure 4.c.

COMPRESd thermal wavfrom a pair oa frequency scorrelation imustrates the de

G , which facilit previous sectpression side rength of the sn thermograph

through pulse

0,3,5,6i =

e 3.a, using thure 4.b). It is oased detection

. SNR of defec

SSION ve imaging, exof halogen lamsweep of 0.0

mage extractedefect detection

tates enhancedtion 2 & threelobes in pulse

side lobes, thishy compatiblee compression

(5

his stimulationobserved from

n, especially in

ts.

xperimentationmps with peak1Hz to 0.1Hzd from variousn capability o

d e e s e n

)

n m n

n k z, s f

Proc. of SPIE Vol. 8705 870510-4

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/23/2013 Terms of Use: http://spiedl.org/terms

OO

OO

OO

WW

WW

WA

NA

O)

CO

Nor

mal

ized

cor

rela

tion

coef

ficie

nt 0.4

0.4

0.4

0.4

0.3

0.3

0.3

FMTWI I

Bar

-15.

I Barker

0.6

0.55

0.5

0.45

0.4

0.35

In pulse comprofiles. Thesin side lobes i

Generally pulgiven by

Figure 6 illuscapability of

Defect detectdetection of bconsiderationbased process

[1] AlmApp

[2] BussApp

a Figure 5

mpression, oscse side lobes aincreases the

lse compressi

strates the PSBarker coded

tion capabilitiblind holes lon. Results pressing technique

mond, D. P. anpl. Phys., 27, 1se, G., Wu, D

pl. Phys., 71, 3

5. Correlation

cillations in thand the main lmain lobe ene

Figure 6.. Pea

ion of thermal

SL of a non dstimulation a

ies of variousocated at diffesented based oes provides be

nd Lau, S. K., 1063-1069 (19D. and Karpen3962 (1992).

b

image of a. FM

he spectrum olobe of the coergy and conse

ak to side lobe l

l profiles amo

PSL 20 lo= ∗

defective pixeand its enhance

6. Cs excitation srent depths. A

on the proposeetter defect de

RE“Defect sizin

994). , W., “Therm

MTWI at 8.4 s

of thermal resompression proequently impr

level ratio of FM

ong these sche

Sidelobe og Correlation⎛⎜⎝

el profile amoed detection c

CONCLUSIchemes have

A quantitativeed non-station

etection capab

REFERENCEng by transient

mal wave imag

b. DFMTWI a

sponse are conofile share theroves detectio

MTWI, DFMTW

emes has been

magnituden peak value

⎞⎠

ong these schecapability in P

IONS been studied

e comparison hnary excitationilities than the

ES t thermograph

ging with phas

c

at7.5s c. Barker

ntributing for e total energy.n sensitivity w

WI and Barker.

n compared u

dB⎞⎟⎠

emes, proved PC.

d on carbon fhas been carrin schemes shoe conventiona

hy I-An analyt

se sensitive m

c er at 4.2s.

r side lobes in. So the reducwith PC system

using peak to s

the side lobe

fiber reinforceried out by takows, that pulsal phase based

tical treatmen

modulated ther

n compressionction of energym.

side lobe ratio

(6

e compression

ed plastics foking SNR intoe compressionanalysis.

nt,” J. Phys. D

rmography,” J

n y

o

)

n

r o n

D:

J.

Proc. of SPIE Vol. 8705 870510-5

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/23/2013 Terms of Use: http://spiedl.org/terms

[3] Mulaveesala, R. and Tuli, S., “Implementation of frequency-modulated thermal wave imaging for non-destructive sub-surface defect detection,” Insight: Non-Destructive Testing and Condition Monitoring., 47 (4), 206-208 (2005).

[4] Mulaveesala, R. and Tuli, S., “Theory of frequency modulated thermal wave imaging for non-destructive sub-surface defect detection,” Appl. Phys. Lett., 89, 191913 (2006).

[5] Maldague, X. and Marinetti, S., “Pulsed phase thermography,” J. Appl. Phys., 79, 2694-2698 (1996). [6] Mulaveesala, R., Somayajulu, V. J. and Pushpraj, S., “Pulse compression approach to infrared nondestructive

characterization,” Rev. Sci. Instrum., 79 (9), 094901 (2008). [7] Tabatabaei, N. and Mandelis, A., “Thermal wave radar: A novel subsurface imaging modality with extended

depth-resolution dynamic range,” Rev. Sci. Instrum., 80, 034902 (2009). [8] Ghali, V. S. and Mulaveesala, R., “Frequency modulated thermal wave imaging techniques for non-destructive

testing,” insight, 52 (9), 475-480 (2010). [9] Mulaveesala, R., Tuli, S. and Pal, P., “Interface study of bonded wafers by digitized linear frequency modulated

thermal wave imaging,” Sensors and Actuators A, 128, 209–216 (2006). [10] Mulaveesala, R. and Ghali, V. S., “Coded excitation for infrared non-destructive testing of carbon fiber

reinforced plastics,” Rev. Sci. Instrum, 82, 054902 (2011). [11] Tabatabaei, N. and Mandelis, A., “Thermal coherence tomography using match filter binary phase coded

diffusion waves,” Phys. Rev. Lett. 107, 165901 (2011). [12] Mulaveesala, R., Ghali, V. S. and Arora, V., “Applications of non stationary thermal wave imaging for

characterization of CFRP,” Electronics letters, 49 (2), 118 – 119 (2013). [13] Ghali, V. S. and Mulaveesala R., "Comparative data processing approaches for thermal wave imaging

techniques for non-destructive testing," Sensing and Imaging International., DOI 10.1007/s11220-011-0059-0 (2011).

[14] Ghali, V. S. and Mulaveesala, R., “Quadratic frequency modulated thermal Wave imaging for non-destructive testing,” Progress In Electromagnetics Research M, 26, 11-22 (2012).

Proc. of SPIE Vol. 8705 870510-6

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 09/23/2013 Terms of Use: http://spiedl.org/terms