On the Application of Liquid-Crystal Thermographyfor the Nondestructive Detection of Delaminationin Composite Structures

Akbar Afaghi Khatibi, Hadi Rezghi MalekiSchool of Mechanical Engineering, Faculty of Engineering, University of Tehran, 11365-4563 Tehran, Iran

To prevent the catastrophic failure of composite struc-tures, several nondestructive testing methods havebeen developed. In this study, a new technique to eval-uate these structures by using thermochromic liquidcrystal (TLC) thermography is presented. Temperaturegradients produced by sensitive liquid-crystals areused to detect delamination in composite specimens.The effects of constituent materials and delaminationsize/location are investigated. The results from TLCthermography are compared to those from infraredthermography. Finally, advantages/disadvantages ofthe new method are discussed. On the basis of theresults from this study, it could be concluded that theTLC thermography can be used as an inexpensive non-destructive testing method in inspection of compositestructures. POLYM. COMPOS., 29:798–803, 2008. ª 2008Society of Plastics Engineers

INTRODUCTION

Because of their high specific modulus and strength,

composite materials have been widely used in aerospace,

marine, chemical, and electrical/electronic industries as

well as infrastructures and bio applications. Generally

speaking, defects can be produced in any structure and

component during manufacturing, handling, or under serv-

ice conditions. To avoid catastrophic failure of composite

structures, it is necessary to find appropriate inspection

techniques to obtain information about the integrity of

these structures. Nondestructive testing (NDT) can offer a

solution in this regard. Among well known techniques in

NDT field applied to composite structures are radiogra-

phy, thermography, and ultrasound methods [1–6].

Delamination is one of the common defects in compos-

ite components and could be detected by using ultrasound

techniques. However, the necessity of having an interme-

diate material in these techniques is putting a limitation

on their usage. In addition, it is possible for the interme-

diate media to penetrate into the structure and degrade or

even destroy the interfacial bonding between fiber and

matrix 7. Considering radiography, a long term usage of

x- or g-radiation can reduce the life-time of resin 8. In

addition, low contrast of images and high initial invest-

ment for labs and dark rooms are some of the other disad-

vantages of radiography techniques 9. Because of these

limitations, other NDT methods such as low frequency or

thermography techniques are mostly preferred in compos-

ite industries.

In this work, a new method of thermography by using

thermochromic liquid crystals (TLC) to evaluate the in-

tegrity of composite structures was developed. At the

heart of TLC based thermography systems are TLC.

These are materials that change their molecular and opti-

cal properties with temperature 10. The liquid-crystals

change from transparent at low temperatures, to red, yel-

low, green, blue, and violet as the temperature is

increased. At higher temperatures, the liquid-crystals

return to the transparent state. On the basis of the sensi-

tivity of TLC to acting temperature, it is possible to

detect invisible defects including delamination in compos-

ite structures. To investigate the capabilities of this new

method and to study the effect of constitutive materials as

well as size/location of delamination, several samples

using different combination of fibers and resins with arti-

ficial defects were manufactured and appropriate examina-

tions were conducted.

PRINCIPLES OF TLC THERMOGRAPHY

Thermography provides a description of the surface

temperature changes and consequently allows detection of

defects or anomalies in the underlying material because

of changes in the surface temperature. The difference

between thermal coefficients of defects and the surround-

ing materials influences the temperature distribution.

Depending on how the thermal load is supplied to the

sample, there are two types of thermography, the so

called passive and active methods [7, 11]. The passive

approach examines structures, which are naturally at

Correspondence to: Akbar Afaghi Khatibi; e-mail: aafaghi@ut.ac.ir

DOI 10.1002/pc.20453

Published online in Wiley InterScience (www.interscience.wiley.com).

VVC 2008 Society of Plastics Engineers

POLYMER COMPOSITES—-2008

different (often higher) temperature than ambient, while

in the case of the active approach, an external stimulus is

necessary to induce relevant thermal contrasts.

The development of liquid-crystal based thermography

over the past 30 years has provided researchers with a rel-

atively inexpensive technique for visualizing and meas-

uring surface temperature. Scientists have successfully

used TLC thermography to investigate various thermal

phenomena in a wide variety of applications [12–16].

Fundamentally, a liquid-crystal is a thermodynamic phase

that is between the pure solid and pure liquid phases of

matter and exists in some organic compounds under cer-

tain conditions. At temperatures below the liquid-crystal’s

event temperature, a TLC will be in the solid state and

will appear transparent. At its event temperature, the TLC

will reflect a unique wavelength of visible light (i.e.,

color) that can be readily captured by a color camera. As

the temperature rises through the TLC bandwidth, the

reflected color of the TLC will change to red, yellow,

green, blue, and violet. A simple, two color-temperature

designator typically describes this response. For example,

R40C10W, the TLC used in this study, is one of TLC for-

mulation. The R40C signifies that the red-start or event

temperature of the TLC is 408C. The 10W signifies that

the blue-start temperature is at 108C above the red start

temperature and this provides users with a crude estimate

of the formulation’s active bandwidth. These color

changes are repeatable and reversible as long as the TLC

is not physically or chemically damaged. Finally, when

the temperature exceeds the TLC clearing point tempera-

ture, the material will enter the pure liquid state and will

revert back to being transparent 10. TLC formulations

with activation temperatures ranging from �308C to

1208C and bandwidths ranging from 0.58C to 308C are

commercially available.

PREPARATION OF SAMPLES

The most common defect, in composite structures,

which can occur during manufacturing or under service

conditions, is delamination. In this study, samples with ar-

tificial defects to simulate delamination were manufac-

tured. Hand-layup method was used to fabricate compos-

ite laminates containing eight layers. E-glass, carbon and

Kevlar-49 fabric fiber with R440 epoxy, and/or polyester

resin were used in this work. Two layers of Kevlar/epoxy

were used as skin for hybrid composites with extra six

layers of glass/epoxy as their core. The dimensions of

samples with glass fibers as well as hybrid samples were

150 � 150 mm2 while specimens manufactured by carbon

fibers were 100 � 100 mm2. The final thickness of com-

posite laminates with glass and carbon fibers, containing

eight layers, was 4 mm while for hybrid samples with 10

layers, it was 6.5 mm. Using a piece of Mylar1

tape, an

artificial delamination was created in all samples. The

dimensions of artificial defect for glass and hybrid sam-

ples were 25 � 25 mm2 while a 20 � 20 mm2 defect size

was used for the specimens containing carbon fibers. All

samples were then cured at room temperature for 24 hr.

To prepare specimens for TLC thermography, one side

of each sample was coated with a thin layer of black

paint. Using an air brush, R40C10W liquid-crystal was

then sprayed on this side of the specimen with an approx-

imate thickness of 25 mm. An isothermal container was

used to apply a steady-state heat flow to the unpainted

side of the sample, as shown in Fig. 1. Changes on the

surface temperature of test specimens due to the existence

of defects was then recorded using a digital camera.

FIG. 1. Schematic arrangement for LC thermography.

FIG. 2. Thermography results for a sample without any defect (a) glass/epoxy specimen, (b) IR, and (c)

TLC thermography. [Color figure can be viewed in the online issue, which is available at www.interscience.

wiley.com.]

DOI 10.1002/pc POLYMER COMPOSITES—-2008 799

RESULTS AND DISCUSSION

To evaluate the results from TLC thermography, all

specimens were also examined by infrared (IR) thermog-

raphy. Figure 2 shows a typical result from both thermog-

raphy methods for a sample without any defect. It should

be noted that due to the existence of defects and conse-

quent differences in heat coefficients between composite

laminate and delaminated area (air), different behavior

FIG. 3. Effect of defect location (a) the location of defects (b) E-glass/epoxy specimen, (c) IR thermogra-

phy, and (d) TLC thermography results. [Color figure can be viewed in the online issue, which is available

at www.interscience.wiley.com.]

FIG. 4. Detectable size of defects (a) E-glass/epoxy specimen (b) IR

thermography, and (c) and (d) TLC thermography results before and af-

ter image processing, respectively. [Color figure can be viewed in the

online issue, which is available at www.interscience.wiley.com.]

FIG. 5. Effect of laminate thickness on defect detection (a) defect loca-

tion in a specimen with 16 layers of E-glass/epoxy, (b) IR thermography

result, and (c) and (d) TLC thermography before and after image proc-

essing, respectively. [Color figure can be viewed in the online issue,

which is available at www.interscience.wiley.com.]

800 POLYMER COMPOSITES—-2008 DOI 10.1002/pc

was observed by IR and TLC thermography methods. In

the case of IR thermography, since the heated and

scanned surfaces of the laminate are same, therefore, the

surface temperature of delaminated area will be higher

than other parts of the laminate, i.e. the heat coefficient

of air is less than surrounding material. On the other

hand, in TLC thermography, the heated surface is not

same as the scanned one, i.e. the heat flow is coming

through the thickness, as shown in Fig. 1. For this reason,

the temperature of delaminated area should be less than

that of surrounding area without any defect.

To study the capabilities of new TLC thermography,

the effects of constitutive materials including fiber and

resin, laminate configuration, number of layers as well as

the location and size of delamination were studied.

The Effects of Defect Location

E-glass/epoxy samples with two similar artificial

defects which have been located in different locations

were examined by TLC and IR thermography (Fig. 3).

Defect 1 was between second and third layers while

defect 2 was created between forth and fifth layers, as

shown in Fig. 3a. It can be seen from Figs. 3c and d both

defects were detected by TLC and IR thermography.

Since the scanning and heating surfaces for IR thermogra-

phy were the same, defect 2 was closer to the IR camera

than defect 1. On the basis of this fact, the maximum

temperature was detected at defect 2 (Fig. 3c). On the

other hand, since the heat flow passes through the thick-

ness of the sample in TLC thermography, there is no con-

siderable difference in temperature gradient of both

defects created by this method, as shown in Fig. 3d. Com-

paring temperature gradients of both defects from IR ther-

mography, it can be concluded that by increasing the dis-

tance between scanning surface and the plane of defect, it

will be difficult to detect any fault in composite lami-

nates. Regarding TLC thermography, although it would

be difficult to determine the location of the defect by this

method, but it is possible to detect any fault regardless of

the distance between scanning surface and the plane of

defect.

The Effect of Defect Size

To find the minimum detectable size of defects by

TLC thermography, E-glass/epoxy samples with three ar-

tificial defects having dimensions of 10 � 10 mm2, 20 �20 mm2, and 30 � 30 mm2 were prepared. All of these

defects were located at the laminate mid-plane. Figure 4

exhibits the results of IR and TLC thermography of these

specimens. The small defect was not detected by IR ther-

mography at all and also it was difficult to find it through

TLC thermography, before using an image processor.

However, after processing the TLC image using a com-

FIG. 6. Detection of delamination in carbon/epoxy laminates (a) the

defect location, (b) IR thermography result, and (c) and (d) TLC ther-

mography before and after image processing. [Color figure can be

viewed in the online issue, which is available at www.interscience.

wiley.com.]

FIG. 7. Effect of different resin materials on defect detection capability of TLC thermography, (a) the

defect location on a E-glass/polyester specimen and (b) and (c) TLC thermography before and after image

processing. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.

com.]

DOI 10.1002/pc POLYMER COMPOSITES—-2008 801

mercial software, as shown in Fig. 4d, the location of

defect 1 (10 � 10 mm2 defect) as well as the dimensions

of other defects were much clearer compared to the

results from IR thermography.

The Effects of Laminate Thickness

Composite laminates with 16 layers of E-glass/epoxy

having an artificial defect located at the mid-plane were

fabricated. The thickness of samples was 7 mm. All sam-

ples were examined by IR and TLC thermography. Figure

5 shows the temperature gradient results from both meth-

ods. As it can be seen from these figures both IR and

TLC thermography were able to detect the artificial

defect. It should be noted that the time to achieve the

active temperature of TLC and IR thermography on the

scanning surface of these specimens was longer than that

of thin samples.

The Effect of Constituent Materials

Different matrix and fabric materials were used to

study the effect of constituent materials on the perform-

ance of new TLC thermography. Composite laminates

including E-glass/polyester, carbon/epoxy, carbon/polyes-

ter as well as hybrid samples of Kevlar/E-glass/epoxy

with artificial defects were fabricated. A typical TLC ther-

mography results for carbon/epoxy specimens is shown in

Fig. 6. Both methods were able to detect the artificial

defect successfully. Considering TLC thermography

results of other samples, including E-glass/polyester (Fig.

7), carbon/polyester (Fig. 8), and Kevlar/E-glass/epoxy

hybrid samples (Fig. 9) it can be concluded that the new

method is capable of detecting any defects in a wide

range of materials used in this work. It should be noted

that the time required to reach the activation temperature

of the TLC, i.e. 408C in this study, in composite lami-

nates with epoxy and/or polyester resins was different.

This can be attributed to the difference of heat conduction

coefficients for these samples.

On the basis of the observations from this experimental

study, it is possible to list the characteristics of TLC ther-

mography for the evaluation of composite laminates in-

tegrity. Among advantages, its low initial investment as

well as lower ongoing expenses compared to other NDT

techniques can be mentioned. The surface temperature

FIG. 8. Effect of different resin materials on the defect detection capability of TLC thermography, (a) the

defect location on a carbon/polyester specimen and (b) and (c) TLC thermography before and after image

processing. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.

com.]

FIG. 9. Effect of constitutive materials on the defect detection capability of TLC thermography, (a) the

defect location on a hybrid Kevlar/E-glass/epoxy specimen and (b) and (c) TLC thermography before and after

image processing. [Color figure can be viewed in the online issue, which is available at www.interscience.

wiley.com.]

802 POLYMER COMPOSITES—-2008 DOI 10.1002/pc

gradients can be saved using a normal digital camera. In

addition, it is possible to repeat the test as long as the

TLC is not damaged. It should be noted that, it is possible

to calibrate and measure the temperature of different area

using TLC thermography. However, black painting of

specimen surfaces and the necessity of having access to

both side of the laminates are among disadvantages of

TLC thermography. The suitability of this method for the

detection of small cracks as well as the capabilities of

this new method in scanning/heating from one side

requires additional study.

CONCLUSIONS

TLC thermography offers an inexpensive method to

nondestructive evaluation of composite laminates. The

results from this study indicate that this technique can be

successfully applied to the detection of delaminated areas

in composite structures. However, some of the limitations

of this method may restrict its usage in every composite

structure. Among them one can refer to preparation proce-

dure and accessibility of both sides of the composite

structure during the inspection. From economical point of

view, on the other hand, this new method offers an inex-

pensive alternative method for accurate evaluation of

composite laminates structures.

ACKNOWLEDGMENTS

Authors thank Advanced Thermal Solutions, Inc. for

providing thermochromic liquid crystal for this study.

REFERENCES

1. J. Summerscales, Ed., Non-Destructive Testing of Fibre-Reinforced Plastics Composites, Vol. 1, Elsevier Applied

Science, London (1987).

2. J. Summerscales, Ed., Non-Destructive Testing of Fibre-Reinforced Plastics Composites, Vol. 2, Elsevier Applied

Science, Elsevier Applied Science (1990).

3. K. Srinivasan, W.C. Jackson, B.T. Smith, and J.A. Hinkley,

J. Reinforc. Thermoplast. Compos., 11, 1111 (1992).

4. X.E. Gros, SAMPE J., 31, 29 (1995).

5. R. Kazys and L. Svilainis, Ultrasonics, 35, 367 (1997).

6. X.E. Gros and K. Takahashi, in Proceedings of InternationalConference on Advanced Materials, Egypt, 483 (1998).

7. V. Dattoma, R. Marcuccio, C. Pappalettere, and G. Smith,

NDT E Int., 34, 515 (2001).

8. L.G.C. dos Santos and Y. Kawano, Polym. Degrad. Stabil.,44, 27 (1994).

9. M.F.S.F. de Moura and A.T. Marques, Compos. Part A:Appl. Sci. Manuf., 33, 361 (2002).

10. J. Stasiek, A. Stasiek, M. Jewartowski, and M.W. Collins,

Opt. Laser Technol., 38, 243 (2006).

11. D.E. Bray and D. McBride D, ‘‘Nondestructive Testing Tech-nique,’’ John Wiley, New York (1992).

12. D.E. Metzger, R.S. Bunker, and G. Bosch, J. Turbomachi-nery, 113, 52 (1991).

13. V.M. Balasubramaniam and S.K. Sastry, J. Food Eng., 26,219 (1995).

14. D.B.R. Kenning, T. Kono, and M. Wienecke, Exp. Therm.Fluid. Sci., 25, 219 (2001).

15. P.T. Ireland and T.V. Jones, Meas. Sci. Technol., 11, 969(2000).

16. C.R. Smith, D.R. Sabatino, and T.J. Praisner, Exp. Fluids,30, 190 (2001).

DOI 10.1002/pc POLYMER COMPOSITES—-2008 803