Damage Study Of Cfrp Composites Using Infrared Thermography

DAMAGE STUDY OF CFRP COMPOSITES USING INFRARED THERMOGRAPHY

Sukanta Das

A Thesis Submitted to

Indian Institute of Technology Hyderabad

In Partial Fulfillment of the Requirements for

The Degree of Master of Technology

Department of Mechanical and Aerospace Engineering

June 2016

Copyright © 2016 by Sukanta Das

Declaration

I declare that this written submission represents my ideas in my own words, and where others’ ideas or words have been included, I have adequately cited and referenced the original sources. I also declare that I have adhered to all principles of academic honesty and integrity and have not misrepresented or fabricated or falsified any idea/data /fact/source in my submission. I understand that any violation of the above will be a cause for disciplinary action by the Institute and can also evoke penal action from the sources that have thus not been properly cited, or from whom proper permission has not been taken when needed.

_________________________

(Signature)

_________________________

(SUKANTA DAS)

_________________________

(Roll No)

Approval Sheet

This thesis entitled “Damage study of CFRP composites using Infrared Thermography” by Sukanta Das is approved for the degree of Master of Technology from Indian Institute of Technology (IIT) Hyderabad.

_______________________

(Dr. Viswanath Chinthapenta, Asst. professor) Examiner
Dept. of Mechanical and Aerospace Engineering

_______________________

(Dr. Syed Nizamuddin Khaderi, Asst. professor) Examiner
Dept. of Mechanical and Aerospace Engineering

_______________________

(Dr. Gangadharan Raju, Asst. professor) Adviser
Dept. of Mechanical and Aerospace Engineering

_______________________

(Dr. S. Suriya Prakash, Assoc. professor) Chairman
Dept. of Civil Engineering

Acknowledgements

Many people serve me during the apogee of this work. I am eternally grateful to all of them! Foremost, I like to convey my deep sense of gratitude to my supervisor Prof. (Asst.) Gangadharan Raju for his understanding, guidance and funding during all the phases of my investigation. I am also showing my gratitude to Prof. (Assoc.) M. Ramji and Prof. (Assoc.) Suryakumar S, whose kindly accepted my request to access the Material Characterization Laboratory and Metrology Laboratory respectively for this work. I am especially grateful to our newborn Thermography team: Swaraj Kumar, Naresh Reddy, and Divya Selvaraj (visiting from NIT Trichy), they all provided me with useful inputs that enriched my understanding and enthusiasm along the subject. I too recognize the invaluable assistance of the technical staff in the laboratory, Karthikeyan, A. Praveen, and Pramod Lokhare. I would wish to convey my earnest gratitude to my family in law, special thanks go to my family in West Bengal: my parents Gouranga Sundar Das and Kalpana Das. I thank them ALL for being a role model for me, as persons, as parents and as professionals and for paying me their unconditional support since the start of my journey in IIT Hyderabad and all through my work with their constant love and encouragement, from near and from far. Last but not the least, this work would not be possible without the support of my wife. Tina, I am thankful for all your patience and understanding, your love and sustenance.

Dedicated to

My beloved family

Abstract

Structural integrity evaluation of composite structures requires a comprehensive understanding of the various failure mechanisms like matrix cracking, fiber breakage, fiber-matrix interface failure, delamination and their evolution across micro to macro scales. Non-destructive testing and evaluation (NDT&E) techniques apply to composite structures for discerning the various damage mechanisms and supply data for numerical life estimation tools. Among the NDT methods, infrared thermography is one of the popular non-invasive technique suited for real-time monitoring and gives in-situ information regarding the onset of damages and its development.

In this study, the objective is to investigate damage initiation and progression in carbon fiber reinforced plastic (CFRP) composite under static loading condition. The passive thermography (PT) and active thermography techniques are utilized in damage growth studies in CFRP laminate using an infrared camera. The PT experiments were held away to study the damage information in the thermograms under monotonic ramp loading. Image processing of the thermal images obtained using thermography is used to obtain qualitative information about the damage evolution in CFRP specimens.

Nomenclature

AoI : Area of Interest

CTE : Coefficient of Thermal Expansion
CFRP : Carbon Fiber Reinforced Plastic
EM : Electromagnetic
FPA : Focal Plane Array
FFT : Fast Fourier Transform
FT : Fourier Transform
GFRP : Glass Fiber Reinforced Plastic
IR : Infrared
IRT : Infrared Thermography
LT : Lock-In Thermography
NDT : Non Destructive Testing
NDT&E : Non-Destructive Testing and Evaluation
NIR : Near Infrared
PPT : Pulsed Phase Thermography
PT : Passive Thermography
SNR : Signal-to-Noise Ratio

TT : Transient Thermography
TSR : Thermographic Signal Reconstruction

UD : Unidirectional
Table of Contents

Declaration ii

Approval Sheet iii

Acknowledgements iv

Abstract vi

Nomenclature vii

List of Figures

Figure 1: Sir William Herschel's "calorific rays" experiment [12] 4

Figure 2: Spectral radiation of a blackbody [17]. 5

Figure 3: IR Thermography system for NDT&E [18] 6

Figure 4: Elements to consider in a thermography inspection scenario [19]. 7

Figure 5: Test specimen draft 9

Figure 6: NDT test setup. (a) passive thermography (b) active thermography for LT & TT 10

Figure 7: Temperature and stress variation plot with respect to time for specimen 03 12

Figure 8: Linear variation of stress and temperature in first phase (Thermoelastic effect). 13

Figure 9: Maximum normalized subtracted temperature timing graph for specimen 01 13

: Normalized subtracted thermogram with damage indications for specimen 01 at different time (t). 14

Figure 11: Normalized subtracted thermogram with damage indications for specimen 02 at different time (t). 14

Figure 12: Normalized subtracted thermogram with damage indications for specimen 03 at different time (t). 14

Figure 13 : Phase diagram and temperature variation of point (1) and (2) for specimen 01. (a) Phase plot using lock-in thermogram, (b) Zoomed part of the damage location from the (a), (c) Temperature variation of point (1) and (2) over time. 15

Figure 14: Microscopic images. (a) Damage 01 and 02 after image stitching adjacent to damage image (b) and (c). (b) Microscopic image of damage 01 at 10x zoom for section 1 of specimen 01. (c) Microscopic image of damage 02 at 10x zoom for section 1 of specimen 01. 16

Figure 15: Test specimen with open hole at center. 18

Figure 16: Stress – Strain diagram for open hole CFRP specimen 1 and 2. 20

Figure 17: Temperature – Stress diagram for open hole specimen 1 and 2 21

Figure 18: Normalized temperature of different AoIs after subtracting from health AoI 21

Figure 19: Maximum normalized subtracted temperature timing graph for open hole specimen 01 22

Figure 20: Open hole specimen 01 Thermograms. (a) Raw, (b) Normalized, (c) Projected normalized-subtracted thermograms of the damage frames. 22

Figure 21: Phase thermogram of at different frequencies, (a) Specimen 1, (b) Specimen 2 and (c) Healthy specimen 23

Figure 22: Microscopic image of section 03 (near to hole), through thickness matrix crack. 23

Figure 23: Microscopic image of section 01 (far away from hole), through thickness matrix crack. 24

Figure 24: Microscopic image of section 02 (away from hole), through thickness matrix crack. 24

Figure 25: The electromagnetic spectrum [18]. 29

Figure 26: Periodic thermal wave after reaching a solid barrier for the low frequency wave [20]. 30

Introduction

Background

Like every new developed technic, Nondestructive testing (NDT) and evaluation also features a wide, diverse, and interesting narration. At that spot where many pioneers who discovered individual methods and many more who developed and improved those methods out of necessity. The bending point and the speedy development of NDT were boosted near after the World War II. Over the period of developments in NDT, ultrasonic, X-rays, liquid penetration, magnetic partials, and eddy currents become the giant tests of NDT. Numerous applications of these techniques are found from the production lines to the material characterization under more controlled environments, schedule maintenance of structural components, etc. Nondestructive testing and evaluation techniques (NDT&E) are an invaluable inspection tool, though methods are required to be reliable, economical, sensitive, user-friendly and fast [1].

A few centuries later, in 1800, Herschel discovered the existence of infrared (IR) rays through a famous experiment at Royal Society of London. Since after the discovery of IR rays, most of its applications and research were done for military applications. After several development IR detectors, in the 1960s and 1970s, the first commercial IR camera came into the market. The availability of commercial IR cameras, emerging from unclassified military technology gave a new direction to NDT applications. Infrared Thermography, among the different NDT&E techniques, stands one of the popular non-invasive technique suited for real-time monitoring whenever a thermal contrast between the area of interest (AoI), commonly damage zone and the background, commonly health zone. If the background and AoI are in thermal equilibrium, an external stimulator can be used to create a temperature difference between them, which called as active thermography [2].

Literature Survey

C. Colombo F et. al. studied to understand and predict fatigue behavior of a GFRP composite [], from thermographic observations. They studied the damage evolution under four loading conditions namely static, interrupted static, stepwise dynamic and fatigue tests. In static loading, they were able to identify the damage stress considering the thermoelastic effect. Later, this damage stress correlate with the fatigue behavior of the material [3]. Similarly, to C. Colombo’s work, W. Harizi et. al. study on damage characterization of GFRP composite [] under static loading using passive thermography. They were able to capture the thermoelastic effect of GFRP and using this principle they were able to define the damage stress [4]. Similarly, J. M. Roche et. al. also studied in situ damage detection and monitoring in GFRP woven composite [], during a mechanical testing. During monotonic tensile loading, damages were monitor by passive thermography, and intermediate pauses, i.e. a constant stress in which pulse thermography was done to evaluate the possibility to detect the damages generated during the mechanical testing. Both active and passive thermography were able to identify the damages [5].

Interesting, other than the above-mentioned works, a good number of researchers were worked on damage study in GFRP composite under monotonic ramp loading [6][7][8], and very few were worked on damage study in unidirectional CFRP composite under static loading [9][10].

Research Objective

The main objective of this work is to:

“Investigate damage initiation and progression in unidirectional carbon fiber reinforced plastic composite under static loading condition using infrared thermography.”

In order to achieve the objective of the work, a series of specific objectives are required and can be stated as, to:

Review the fundamental concepts behind Thermography (Chapter 1)

Review the fundamental principles behind passive thermography, data acquisition, principle of thermoelastic effect and image processing techniques for qualitative analysis (Chapter 2)

Review the fundamental principles behind active thermography, i.e. Lock-in and pulse phase thermography for qualitative analysis (Chapter 3)

Organization

This thesis is organized into six main chapters. In Chapter 1, the place of infrared thermography in the NDT&E scene is first established. Some basic theory of infrared thermography (IRT), and a different type of IRT techniques discussed. Experimental concepts such as data acquisition, thermogram sequence and defect detection techniques by passive thermography are discussed as an introduction to the basic theory behind passive thermography, offered in Chapter 2. Later, two fundamental concepts are then carefully reviewed, namely the principle of Thermal waves generation due to modulated heat and principle of Thermoelastic effect due to an external load in this chapter. Chapter 3, discusses the active thermography technique, followed by a description of how Lock-in thermography (LT) and Pulse Phase Thermography (PPT) worked. At the end, last three chapters, chapter 4, 5 and 6, were discuss three case study namely; “Study of Lock-in and Transient thermography on CFRP composite with different depth of holes”; “Infrared thermography technics for damage study on unidirectional CFRP under static load” and Damage study on unidirectional CFRP composite with an open hole under static loading.

Infrared Thermography

Short history of the Infrared

Even before Max Planck published his theory of radiation on 14 December 1900, measurement of temperature was of concern to Mankind. The first law of thermodynamics introduces the energy conservation principle and explains that any (industrial) process consuming energy will assure a great part of this energy be transformed into heat (following the law of entropy). Temperature is, therefore, an important parameter, to quantify. The glass thermometer, invented by Galileo in 1593, was the first instrument for quantitative temperature measurement (Wise 1988). It allowed Herschel in 1800 to discover the infrared spectrum [11].

The German astronomer, W. Herschel, in 1800 discovered the existence of infrared rays during his experiment with a glass prism and thermometers (Figure 1). While moving the thermometers, he evaluated the temperature in each color zone. He found that the temperature was increasing from the violet to the reddish portion of the spectrum. When he moved the thermometer beyond the red part, he found that this area experienced had the highest temperature. Thus, he made a big breakthrough and this part of the spectrum was named "calorific rays". Today, calorific rays recognize as “infrared” radiation. Hence, infrared radiation is electromagnetic radiation covering the spectral region between the red end of visible light and microwaves [12].

Planck’s Law

Planck’s law is one of the most important laws governing the thermal emission, and thermal emission is too an important parameter for IRT. According to this law, the distribution of emitted energy is a function of the wavelength for a given temperature, i.e. for a given temperature the magnitude of the emitted radiation varies with wavelength. Figure 2, shows the magnitude of the radiation varies with wavelength [1].

Where,

h – Planck's constant (6.626076 x l0-34 J.s);

C – Speed of light (~3 x 108 m.s-1);

K – Boltzmann's constant (1.381 x l0-23 J.K-1).

Emissivity

Emissivity is a material property that states the ability to emit energy from the surface. It is expressed as the ratio of the radiation emitted by a surface to the radiation emitted by the blackbody in the same conditions of temperature, direction and spectral band of interest. Emissivity is a unitless quantity and spans from 0 to 1. Emissivity depends on the surface orientation, temperature, and wavelength. Surfaces with a low emissivity (polished metals) act as a mirror. It is then difficult to measure the actual temperature of a sample surface because of the influence of emitted radiations from the surrounding objects. There are various ways to dilute the impeachment of the environmental reflections. One of them is to cover the inspected surface with a high emissivity flat paint (with ) [1].

Infrared thermography system

An infrared thermography system consists of mainly, a thermal excitation source; a target object; a radiometer (IR camera); a signal and image analysis system (PC); and the resulting thermogram (Display). Figure 3 represent the typical IR thermography system.

An optical system collects the emitted infrared energy from a focused object onto the detector. The detector converts this energy into an electrical signal which then is processed and displayed as a value of measured temperature.

Out of two type of detector namely thermal and photonic, thermal detectors are more popular in the industries. Thermal detectors are based on the temperature dependent phenomena (temperature change, thermoelectric effect, etc.). They are practically independent on the wavelength, the element does not need to be cooled and their price is comparatively low [1].

Infrared Thermography in the NDT&E

Non-destructive Testing and Evaluation (NDT&E) involves all inspecting techniques used to investigate a part or material or system without damaging it. The objective of an NDT&E technique is to provide information about flaws and separations; structure; dimensions and metrology; physical and mechanical properties; composition and chemical analysis; stress and dynamic response; signature analysis; and abnormal sources of heat. A wide variety of NDT&E techniques are available, namely, Visual Testing, Ultrasonic Testing, Radiographic Testing, Eddy Current Testing, Magnetic Particle Testing, Liquid Penetrant Testing, and Infrared Thermography; but none of which is able to reveal all the required information. The appropriate technique depends on the thickness and nature of the material being inspected, as well as in the type of discontinuity that must be detected.

Infrared and Thermal testing involve temperature and heat flow measurements to predict or diagnose the failure. IRT is a nondestructive, non-contact and non-intrusive mapping of thermal patterns on the surface of the objects [1]. It can inspect of large areas in a fast and safe manner. Infrared thermography can be divided into two categories based on the temperature difference generated to obtain thermograms: passive thermography and active thermography.

Passive thermography

In passive thermography techniques, the features of interest are naturally at a higher or lower temperature than the background, i.e. is can used for the objects which are naturally in the higher or lower temperature than ambient. In passive thermography, an abnormal temperature profile indicates a potential problem. This abnormality refers to temperature difference with respect to a reference, often refer as thermal contrast. The measured surface temperature can be related to specific behavior or subsurface flaws [1].

Active thermography

In active thermography, an external energy source requires inspecting specimen in order to obtain significant temperature difference witnessing the presence of subsurface flaws. Various test procedures are there namely, pulsed thermography, step heating, lock-in thermography, and vibrothermography. Figure 4 presents a summarized diagram of the different elements to take into account when designing an NDT inspection [1].

Infrared thermography for damage study in unidirectional CFRP composites under static load

Objective

In this work, the aim is to investigate damage initiation and progression in unidirectional carbon fiber reinforced plastic (CFRP) composite under static loading condition. The passive thermography (PT) and active thermography techniques are employed in damage growth studies in CFRP laminate using an infrared camera. The PT experiments were carried out to examine the damage information in the thermograms under monotonic ramp loading. Image processing of the thermal images obtained using thermography is used to get qualitative information about the damage evolution in unidirectional CFRP specimens.

Specimen preparation

The material used in this study is a carbon/epoxy laminate with a 60 % volume fiber content. The test coupons were cut manually from a laminated plate comprising of eight layers of unidirectional dry woven fabric with 210 gm/m2 areal weight. An epoxy resin of Araldite CY230 mixed with the hardener of Araldite HY951 in a ratio of 10:1 was used to impregnate the fabric. The laminates were manufactured using hand lay-up followed by vacuum bagging to improve the fiber volume fraction. The stacking sequence is [0°]8 to get a final specimen thickness (t) of 2.1 mm after curing at room temperature of 28.5 oC, under a vacuum pressure of – 87 kPa for 24 hours. The specimens were endowed with 2 mm thick aluminum (tt) adhesively bonded tabs, ensuring the shear forces transferred through the adhesive to minimize any stress concentration due to the gripper pressure (700 psi). The schematic of the test specimen with dimension is shown Figure 5. For preparing the test coupon, carbide diamond coated milling cutter of diameter 6 mm was used to reduce the machining effects at the edges of coupon.

Test setup

The tensile test was carried out on unidirectional laminates under displacement control mode in MTS Landmark 370, having a maximum load capacity of 100 kN. The micro cracks, delamination, fiber damages and also other damages were observed and recorded using infrared (IR) camera. The IR thermal camera from FLIR SC 5000, equipped with a cooled Indium Antimonide detector allows the smallest of temperature differences to be seen. It is working in waveband range of 2.5 to 5.1 μm and produces the thermal images of 320 x 256 pixels. The camera has Focal Plane Array (FPA) InSb sensor with a high capacity of 25 mK thermal sensitivity[13].

For passive thermogram recording, the IR camera was placed approximately 510 mm far from the surface of the specimen loaded into the MTS machine. The experimental test setups are shown in Figure 6. The camera was connected to a computer for recording the thermogram data. The recording and post-processing of the data were done using Altair software. This software identifies the damaged zones and performs a pixel by pixel analysis of the temperature data. To reduce the environmental noise in the recorded thermogram, whole experimental setup (i.e. MTS machine and the IR camera) was covered with a non-transparent plastic cover. The test room was maintained at a constant temperature of 23 °C. For inspection of the damaged specimens, active thermography adopted using the same camera additionally used two halogen lamps having a maximum capacity of 2.5 kW for the heat wave. For all the active thermography test, lamps and camera have kept a distance of 700 mm and 550 mm from the specimen respectively. All the active thermography tests were conducted using IrNDT software with lock-in and transient thermography module, supported by Automation Technology.

For a microscopic analysis of the pre-identified damage sections from the normalized-subtracted thermogram, Olympus STM6 optical microscope with the objective lens ‘MPlanFL N 10x /0.30’ having a capacity of 10x magnification with a numerical aperture value of 0.3 was used.

Experimental tests

Static loading

Static monotonic tensile tests were performed in the displacement control mode with a crosshead speed of 1 mm/min. The three specimens were loaded on an MTS universal testing machine and during each tensile test, the surface thermogram of the specimens was recorded by the thermal camera. The aim was to capture any onset of micro damages indication in the temperature trend and correlate it with the material under load. Three specimens were loaded up to stress level of 500, 700 and 900 MPa, respectively, and the temperature data acquisition frequency was set at 120, 85 and 55 Hz.

Active Thermography

After tensile loading, every loaded specimen was inspected through active (i.e. Lock-in and transient pulse techniques) thermography to capture the micro and macro damages. A different thermal excitation frequency of 2, 1, 0.34 and 0.2 Hz with corresponding thermal periods 5 for 1st two and 4 for other frequencies, was selected for the lock-in thermography. For transient pulse thermography, a rectangular pulse width of 1.5, 2.25 and 3 sec. with a period time of 10, 15 and 20 sec. respectively was adopted. The recorded thermograms were evaluated using pulse phase techniques to plot the phase diagram. The image acquisition frequency of 300 Hz was chosen for both the techniques. The aim of these tests was to plot the phase diagrams of the loaded specimen and correlate phase delay response of the damage location with the passive thermogram indications.

Microscopic Analysis

For the validation of the thermogram techniques, the specimens were cut corresponding to the critical thermogram zones and inspected under the optical microscope to capture the micro – macro damages. The critical sections of the specimen were cut using carbide diamond coated milling cutter of diameter 3 mm. To reduce machining defects, a different set of cutting parameters, i.e. cutting speed and depth of cut were chosen and applied to cut a healthy specimen. After analyzing all the microscopic images of different cut sections of the healthy specimen corresponding to the different cutting parameter, it was concluded that the least number of defects was observed for the cutting speed of 1000 RPM with the 0.5 mm depth of cut. These optimum parameters were used to cut the different loaded specimens for microscopic analysis.

Test results and discussions

The stress-strain results of the static tests showed an approximately linear behavior at the beginning of the test. Later, the stress in CFRP specimen started to increase slowly, due to different damage mechanisms until it reached the ultimate strength () of 1.175 GPa. By looking at the recorded thermogram, it was very difficult to locate the damaged zones at low stress levels and image processing was done to enhance the hottest zone. The normalizing followed by the subtraction operation between current frame and the previous frame of the recorded film were done with Altair software [9]. Also, during the tests, flashes of rapid temperature change with time was observed in certain regions in the specimens. These flashes were oriented along the fibers indicating the energy release due to fibre breakage and also debonding at the fibre-matrix interface.

Temperature evaluation within the specimen, due to the external load can be considered as an important parameter for the damage state of the material. It should be stressed that in the passive thermogram analysis, the temperature was averaged over an area of interest (AoI) of the specimen. This area corresponds to the central part of the scanned surface of the specimen; avoiding the upper and lower parts of the local influence due to the grips’ temperature, and the external borders for the edge effect due to machining. It is interestingly found that the selection of the AoI has no influence on the results of the temperature trends. The temperatures were evaluated at different AoI and the trends were found to be similar; the only change is the shifting on the temperature scale, depending on the local temperature of the AoI.

All the tested specimens showed a similar temperature variation with time under tensile loading condition. The temperature profile can be schematized in two phases. The first phase characterized by an approximately linear increase of the temperature variation with time. This increase of temperature observed in CFRP specimens is in agreement with the thermoelastic effect () due to the negative coefficient of thermal expansion (CTE) in the fibre direction. The second phase characterized a nonlinear variation in the temperature with time. This can be attributed to the initiation and growth of local micro-damages arising from the pre-existing manufacturing defects [6][4][14]. These damages can be confirmed by micrographic analyses using an optical microscope.

The temperature and the stress trends are represented together in the same graph, given in Figure 7. Temperature-time data were then compared to stress-time data to evaluate the change in the mechanical behavior of the specimens. The first phase is characterized by a linear trend (Figure 8), which can be described with an equation obtained by means of a linear regression analysis. The end of the first zone might be correlated to the end of the linear thermoelastic behavior of the material and beginning of the micro-damages in the material. Hence, the stress value corresponds to the end of the linear thermoelastic phase, consider as damage initiation stress. The average value, obtained from static tests, was equal to 221 MPa and represents the damage initiation stress point () for the unidirectional CFRP specimens [6][9].

Furthermore, to find the location of the damages, image processing was done with the recorded thermogram. The recorded thermogram was first normalized by the first frame of the film and further subtraction was done between two subsequence frames of the normalized film to capture the small temperature rise within the specimen [9]. These arithmetic operations have used Altair software and the plot of the maximum normalized-subtracted temperature of the AoI versus time is shown in Figure 9. A threshold value of 0.006 of the “maximum normalized-subtracted temperature” was chosen for identifying the damages in the specimen. By using the threshold value, the critical images of CFRP specimens showing the evolution of damages with time for different stress levels (500 MPa, 700 MPa, 900 MPa) are represented in Figure 11, Figure 10, and Figure 12.

For active thermography, both lock-in and transient pulse thermography were done for specimen 01 and 02. For specimen 01, lock-in thermography able to capture the damages, but for specimen 02, both the techniques were not able to capture the damages for the selected frequency range. The lock-in test result showed Figure 13 (a) represent the phase information on loaded specimen 01 (500 MPa) and healthy specimen. It was clearly identified from the phase diagram of the loaded specimen compare to a healthy specimen that, damages are located a top portion of the specimen 01. After pixel by pixel analysis of this phase value, we could summarize that the damages location of the specimen 01 is nearly same with passive thermogram damage indications. The temperature variation of the damage point (1) and health point (2) corresponds to loaded specimen 01 and healthy specimen were shown in Figure 13 (c). The deviation in cooling profile between healthy and loaded specimen indicates the damage present at the point (1). The above analysis clearly indicates the damages for the specimen 01 and agree with respective passive thermogram damage indications.

To validate passive thermogram damage indications, specimens were cut along the damage indications and inspected under a microscope. For specimen 01, a list of the damages and their corresponding dimensions were mentioned in Table 1. The microscopic image (Figure 14) of critical section 1, corresponding to specimen 1, was able to highlight local delamination, which was also indicated by the passive thermogram.

Table 1 : Damage dimensions from microscopic images for specimen 01

Conclusions

Quasi-static tensile experiments were carried out unidirectional CFRP specimens to study the damage initiation and progression using thermography. Both passive and active thermography techniques were used to monitor the damage progression in unidirectional laminates. Passive thermography results were able to capture the thermoelastic effect showing a temperature increase with loading within the elastic limit. Also, using PT we were able to identify the damage initiation stress point (approximately 221 MPa) beyond which the temperature variation becomes nonlinear with load showing the growth of damage in CFRP specimens. Further image processing operations were performed on thermograms acquired using PT and a threshold value was adopted to identify the damages in the CFRP specimens. Active thermography techniques, namely, lock-in and transient pulse thermography were then used to validate the damage identified using PT. Finally, the specimens were cut at the damage location identified by PT and the micrographic images were captured using an optical microscope. In addition, an experimental study was carried to determine the optimal cutting speed and depth of cut to minimize the machining damage. The microscopic images obtained at the damaged section agreed with the results obtained by PT.

Damage study on unidirectional CFRP composites with open hole under static load using infrared thermography

Objective

Similar to previous work, the aim of this work is to investigate damage initiation and progression in unidirectional carbon fiber reinforced plastic composite with an open hole under static loading condition, using thermography techniques.

Specimen preparation

The material used in this study is same as described in the previous problem. Additionally, a through hole drilled in the center of the specimens. The diameter of the hole is 10 mm and the specimen thickness is 2 mm (Figure 15).

Test setup

The test setups and environmental conditions were kept same as described in the previous study. For passive thermography, the space between, IR camera and specimen loaded into the MTS machine were placed approximately 540 mm and for active thermography, lamps and camera have kept a distance of 1100 mm and 950 mm from the specimen respectively.

Experimental tests

Static loading

The two specimens were loaded up to stress level of 445 and 415 MPa respectively on an MTS universal testing machine. During each tensile test, the surface thermogram of the specimens was recorded by the thermal camera. Standardized to the earlier test, here also the aim was to capture any onset of micro damages indication in the temperature trend and correlate it with the material under load. The temperature and stress-strain data acquisition frequency for the specimen 1 and 2 was set at 80 and 70 Hz respectively.

Active Thermography

A different thermal excitation frequency of 0.562, 0.2, 0.15, 0.1, 0.05, 0.032, 0.0216, 0.015, 0.0122 and 0.01104 Hz with corresponding thermal periods of five for 1st two frequencies and three for the rest of the frequencies was selected for the lock-in thermography. The image acquisition frequency of 350, 290, 290, 195, 95, 60, 40, 28, 23 and 20 Hz respectively, was chosen for LT techniques.

Microscopic Analysis

Microscopic analysis was done exactly described in the previous problem.

Test results and discussions

The stress-strain results agree on the previous test results. The stress-strain plot showed an approximately linear behavior at the beginning of the test. Afterward, the stress in CFRP specimen started to increase slowly, due to different damage mechanisms until it reached the ultimate strength (Figure 16). By looking at the recorded thermogram, this time, it was less difficult to identify the damaged of low-stress levels, but, withal it was difficult to see the damage initiation. Standardized image processing was performed to raise the damage initiation zone. From normalized film, the mean temperature of different AoI was plotted and found that some of the AoI (near to the hole) shows ideal until it stress reaches approximately 213 MPa., after this point a gradual increase in temperature found, and later these gradual increase in temperature converted in sudden jump in temperature at higher stress level. For, deeper study of this temperature behavior, some of the critical AoI (containing gradual increase phenomena) choose and subtracted with a mean temperature of approximately healthy AoI. From the healthy subtracted plot, it can be concluded that, temperature increase rate within the stress level 0 MPa to 200 MPa was almost negligible, but after 200 MPa to 350 MPa, temperature increase at a faster rate [i.e. ] for next 1 sec. and slowdown again to accumulate as pre-exiting micro damages,. Subsequently, this rate change to and for a fraction of time (i.e. 0.47 sec. and 0.05 sec.) at higher stress level, which can be linked up to fiber damage.

As reported in the previous problem, AoI selected at the center part of the specimen nearer to the hole chosen for thermoelastic effect study. Standardized to the premature problem, in this work also we found that exact of the AoI has no influence on the result of the temperature trends. Here too, the temperature profile can be schematized in two phases. Applying the thermoelastic effect concept, it is found that damage starts initiated at ~ 170 MPa. The temperature and the stress trends are represented together in the same graph, given in Figure 17. Furthermore, to find the location of the damages, normalized and normalized-subtracted image processing was done with the recorded thermogram. Standardized to the earlier test, threshold value techniques applied to normalized-subtracted thermogram, and able to capture the micro – macro damages. All the frames which are above the threshold level (Figure 19) were closely inspected, and later, all identified damage frames are collected from the Normalized-subtracted film and added to the project, all damage initiation over time on a single image. The final projected image, able to capture approximately to the actual failure pattern of the CFRP specimen. The raw thermogram and project image were shown in Figure 20.

For active thermography, lock-in thermography performed. The phase diagrams of the specimen 01 and 02 with a reference specimen were shown in Figure 21. It is observed that defects of specimen 01 were detected at the highest frequencies of 2.00 Hz, 0.40 Hz, and 0.20 Hz. As the frequency decreased to 0.15 Hz, the contrast begins to reduced and when it goes down to 0.10 Hz, the surface defects in phase images became invisible [15]. Highest phase contrast found similar to the damage pattern of passive thermography. Referable to the limitations, we are not too capable of seeing the deeper damages, and only surface damages are captured with the selected frequencies. From the phase plot, it can be concluding that there is very less significant damage present in specimen 02.

For validating passive thermogram damage indications, here also specimens were cut at a distance of 5, 20 and 55 mm (both side) from the center of the hole and inspected under a microscope. The microscopic images are extensively inspected, and able to summarized that, near the hole section (name as 03) having through-thickness transverse matrix crack (Figure 22) as it was experienced maximum stress during loading. There is some evidence of through thickness matrix crack when to investigate far way hole section (namely 02 and 01), but the damage width is comparatively less (Figure 24 and Figure 23).

Summary

Quasi-static tensile experiments were carried out unidirectional CFRP specimens with an open hole to study the damage initiation and progression using thermography. Both passive and active thermography techniques were used to monitor the damage progression in UD laminates with an open hole. Passive thermography able to identify the damage initiation stress point (approximately 170 MPa) beyond which the temperature variation becomes nonlinear with load showing the growth of damage in CFRP specimens. Further image processing operations were performed on thermograms acquired using PT and a threshold value was adopted to identify the damages in the CFRP specimens. Active thermography technique, lock-in thermography were then used to validate the damage identified using PT. Finally, the specimens were cut at the damage location identified by PT and the micrographic images were captured using an optical microscope. In addition, an experimental study was carried to determine the optimal cutting speed and depth of cut to minimize the machining damage. The microscopic images obtained at the damaged section agreed with the results obtained by PT. In the future, we want to study the damage progression on different CFRP specimens subject to dynamic loading using infrared thermography.

Conclusions

In this study, the advantages and the limitations of infrared thermography were studied. Considering those advantages and limitations, we did the damage initiation studied on two problems. In the first problem, the study was on unidirectional CFRP composites subjected to static loading and the second was a unidirectional CFRP composite with an open hole subjected to a same static load. To establish both the problems, both the cases done in three stages, namely ‘loading and passive thermography’, ‘Active thermography’ lastly, ‘microscopic analysis’.

In both the cases, passive thermography along with advanced image processing was able to identify the damage initiation stress values for unidirectional CFRP composite. For the validation of passive thermography results, the active thermography technique, namely Lock-in thermography adopted, but due to some of the limitations of lock-in thermography, deeper damages were not able to capture in both the cases. At the end of each study, specimens were cut and inspected under microspore to validate the passive thermography results.

Summary of conference paper

This research work has been the source of several scientific communications. The following article was submitted as 4 pages and 6-page conference paper respectively.

[1] S. Das, N. Reddy, and G. Raju, “Damage Evolution Studies in Carbon Fiber Reinforced Polymer Composites using Active and Passive Thermography,” in NDE 2015, 2015, pp. 1–4.

[2] S. Das, N. Reddy, and G. Raju, “Damage growth study in unidirectional CFRP composites using infrared thermography,” in SICE 2016, 2016, pp. 1–6.

References

[1] X. P. V. Maldague, Theory and Practice of Infrared Technology for Nondestructive Testing. Wiley, 2001.

[2] C. Hellier, Handbook of Nondestructive Evaluation, Second. United States: McGraw-Hill Education, 2013.

[3] C. Colombo, L. Vergani, and M. Burman, “Static and fatigue characterisation of new basalt fibre reinforced composites,” Compos. Struct., vol. 94, no. 3, pp. 1165–1174, 2012.

[4] W. Harizi, S. Chaki, G. Bourse, and M. Ourak, “Composites : Part B Mechanical damage assessment of Glass Fiber-Reinforced Polymer composites using passive infrared thermography,” Compos. Part B, vol. 59, pp. 74–79, 2014.

[5] J. M. Roche, B. Lamboul, G. Bai, F. Passilly, A. Mavel, and G. Grail, “Passive and active thermography for in situ damage monitoring in woven composites during mechanical testing,” Quant. Nondestruct. Eval., vol. 562, pp. 555–562, 2012.

[6] C. Colombo, F. Libonati, and L. Vergani, “Fatigue damage in GFRP,” Int. J. Struct. Integr., vol. 3, no. 4, pp. 424–440, 2012.

[7] T. Lisle, C. Bouvet, M. L. Pastor, P. Margueres, and R. Prieto Corral, “Damage analysis and fracture toughness evaluation in a thin woven composite laminate under static tension using infrared thermography,” Compos. Part A Appl. Sci. Manuf., vol. 53, pp. 75–87, 2013.

[8] F. Libonati and L. Vergani, “Damage assessment of composite materials by means of thermographic analyses,” Compos. Part B Eng., vol. 50, pp. 82–90, 2013.

[9] B. Rodriguez, C. Galleguillos, R. Fernández, and F. Lasagni, “Passive Infrared Thermography for Damage Monitoring During Structural Testing of CFRP Parts,” Proc. 16th Eur. Conf. Compos. Mater., no. June, pp. 22–26, 2014.

[10] J. Montesano, Z. Fawaz, and H. Bougherara, “Use of infrared thermography to investigate the fatigue behavior of a carbon fiber reinforced polymer composite,” Compos. Struct., vol. 97, pp. 76–83, 2013.

[11] X. P. V. Maldague, Nondestructive Evaluation of Materials by Infrared Thermography. 1997.

[12] INFRARED PROCESSING AND ANALYSIS CENTER (IPAC), “Herschel Discovers Infrared Light,” IPAC, 2004. [Online]. Available: http://coolcosmos.ipac.caltech.edu/cosmic_classroom/classroom_activities/herschel_bio.html. [Accessed: 20-Jun-2016].

[13] SYSTEM FLIR, “FLIR SC5000 Series,” FLIR® Systems, 2016. [Online]. Available: http://www.flir.co.uk/cs/display/?id=42577. [Accessed: 10-May-2016].

[14] N. I. Baurova, W. Hao, and O. Xiao, “Microstructure of Carbon Fiber and Carbon Reinforced Plastic,” vol. 2013, no. October, pp. 28–32, 2013.

[15] S. Ranjit, M. Choi, and W. Kim, “Quantification of defects depth in glass fiber reinforced plastic plate by infrared lock-in thermography,” J. Mech. Sci. Technol., vol. 30, no. 3, pp. 1111–1118, 2016.

[16] G. Pitarresi and E. A. Patterson, “A review of the general theory of thermoelastic stress,” no. July 2003, pp. 405–417, 2015.

[17] Wikimedia Commons, “BlackbodySpectrum loglog,” Wikimedia Commons, 2006. [Online]. Available: https://commons.wikimedia.org/wiki/File:BlackbodySpectrum_loglog_150dpi_en.png. [Accessed: 22-Jun-2016].

[18] C. Ibarra Castanedo, “Quantitative subsurface defect evaluation by pulsed phase thermography: depth retrieval with the phase,” Université Laval, 2005.

[19] C. Ibarra-Castanedo, J. R. Tarpani, and X. P. V Maldague, “Nondestructive testing with thermography,” Eur. J. Phys., vol. 34, no. September 2015, pp. S91–S109, 2013.

[20] C. Ibarra Castanedo, “Quantitative subsurface defect evaluation by pulsed phase thermography: depth retrieval with the phase,” Université Laval, 2010.

The Electromagnetic Spectrum

Periodic Thermal Waves

The Fourier’s Law one-dimensional solution for a periodic thermal wave propagating through a semi-infinite homogeneous material may be expressed as [1]

where T0 [°C] is the initial change in temperature produced by the heat source, [rad/s] is the modulation frequency ( with f being the frequency in Hz), [m] is the wavelength; and [m] is the diffusion length given by

Where, [m2/s] is the thermal diffusivity, with k [W/m°C] being the thermal conductivity,  [kg/m3] the density,  [J/kg°C] the specific heat; and f the thermal wave modulation frequency. The thermal wavelength is defined as

Figure 26 shows two thermal waves having the same amplitude but different frequencies. Thermal wave propagation through a solid is a strongly damped phenomenon in all cases, however, the high-frequency thermal wave experience a greater decay than the low-frequency wave. Hence, from Eqn. (B.1) to (B.3), it can be observed that after traveling a distance equal to , the thermal wave has already damped to exp (-1) ≈ 0.37 of its initial value, and by a factor of exp(-2) ≈ 1/535.5 after penetrating a distance equal to . These situations are illustrated in Figure 26, the dotted line corresponds to a pure exponential decay.

Furthermore, the propagation speed is also different for the two thermal waves, since it depends on the modulation frequency

Consequently, low-frequency thermal waves penetrate deeper into the material but they do so at lower speeds than high-frequency waves.

Theory of Thermoelastic Stress Analysis (TSA)

Thermoelastic stress analysis is based on small temperature changes of the order of 0.001 °C that occur when a material is subject to a change in elastic strain, and this physical behavior is typically referred to as the thermoelastic effect. The practical exploitation of the thermoelastic effect using infrared technology during the last two decades has stimulated research, leading to new fields of application. There has been particular interest in the use of thermoelastic stress analysis (TSA) to composite materials although an extension of the theory to anisotropic media was first presented by Biot and in the formulation of a higher order theory for the thermoelastic effect, motivated by the need to provide a theoretical explanation for the experimental evidence of the influence of mean stresses on the thermoelastic signal.

Thermodynamics of the elastic continuum

From the thermodynamic point of view, to investigate the mechanical and thermal behavior, the state of an elastic solid can be described by assigning the strain tensor and temperature as state variables, assuming that only small displacements are present.

Constitutive relationships

Constitutive equations describe the relationship existing between stress and strain fields. The most general form of constitutive relations also takes into consideration the effect of thermal and hygroscopic strains and is shown for the case of elasticity by the following general Duhamel-Neumann relation.

Where is the elastic isothermal stiffness tensor of fourth order with 21 independent components, is the second-order thermal expansion tensor, is the reference temperature and is the reference humidity.

A simpler form of equation (C.1), neglecting the hygroscopic terms, is

And is the second-order tensor of the coefficients of thermal expansion. Eqn. (C.2) can also be written in the following way, which is valid for homogeneous isotropic bodies:

Where and are the Lame constants, given by

where is the linear thermal expansion coefficient for isotropic materials.

A slightly different and simpler set of expressions can be written to describe the constitutive law for the case of a plane stress field. In this case equations (C.3) can be rewritten as

By substituting equations (C.4) and (C.5) into the equations (C.6),

From the other relation in equations (C.6), and using equations (C.4) and (C.5) again, the following expression can be written for as a function of and:

And substituting equation (C.8) into equation (C.7) gives

In the same way, an equivalent expression can be obtained for and. Therefore, the stress-strain-temperature relation in the case of a plane stress field can be summarized as follows:

According to the linear representation of the stress-strain-temperature relations for elastic solids, it is possible to assume also the stress tensor as a state variable, together with temperature.

Thus, a state function, i.e. a thermodynamic function describing the state of the system, such as the internal energy u, may be written as

And

Where equation (C.11b) assumes the important characteristic of state functions that infinitesimal increments are exact differentials, i.e. the change in state variables from A to B is independent of path or transformation used to move from A to B.

Now, the first law of thermodynamics can be expressed as

Where all the terms are specific (i.e. per unit mass), is the internal energy and and are the work and heat respectively exchanged between the system and the external surroundings. According to the signs used in equation (C.12), it is implicit that the convention used is to consider the heat q positive when transferred from the external surroundings to the system and the work positive when done on the system by forces external to it.

The second law of thermodynamics can now also be introduced as

Where s is a new state function, called entropy, defined concerning a reversible process using equation (C.13a). It is worth noting that, when considering a more general irreversible process, there is always a dissipation of energy and, in an adiabatic process,, there is always an increase in the entropy of the system.

If a continuum in equilibrium is considered which undergoes deformations in a quasi-static way (such that there are no inertial forces acting on it, and the kinetic energy is constant), and the continuum is assumed to have small deformations compared with the body dimensions, such that any higher-order terms as in equation (C.12) for work are neglected, then, from the principle of virtual work, the work was done by external forces on the system is equal to the strain energy gained by the continuum. Therefore, considering the system represented by an infinitesimal element of unit volume, the work exchanged with the external surroundings is given by the strain energy density:

Where is the small elastic strain tensor, is the stress tensor and the temperature is considered as constant. For application to a unit mass, it is necessary to multiply by the specific volume whereis the density:

Now, another state function is introduced: the free energy, or Helmholtz thermodynamic potential, per unit mass, as

And its differential form

From the first and second laws of thermodynamics [equations (C.12) and (C.13)],

For local reversible changes; therefore, equation (C.16b) becomes

H is a function of strain and temperature, and is an exact differential; therefore,

And, from comparing equation (18) with equation (19),

At the same time, the entropy s per unit mass is a state function, and the following equation can be written

The specific heat per unit mass at zero strain (i.e. for constant-volume transformations) is defined as

Substituting equations (C.20) and (C.22) in equation (C.21) gives

Classical theory of thermoelastic stress analysis

If equations (C.23) which arise from the definition of the specific heat per unit mass are now considered, and the statement of the second law embodied in equation (C.13b) is substituted, then

from which

In equations (C.24) and (C.25) there is a term representing the partial derivative of the stress tensor on temperature. This term can be developed using at this stage the stress-strain-temperature relation [equation (C.3)], which is valid for homogeneous isotropic materials, and assuming that the Lame elastic parameters are independent of temperature. Thus

And is a function of Young’s modulus, Poisson’s ratio and the coefficient of linear expansion as defined in equation (C.5). The product is the first strain invariant or cubic dilatation ; therefore, from equation (C.23),

Integrating and setting at the starting conditions, when and this gives

Considering and expanding the logarithm term into an infinite power series in which only the first term is retained,

For an adiabatic process, then

which is also the relation as it appears in the pioneering work of Biot.

Using the stress-strain-temperature relationship [equation (C.3)] again and the connection between the linear expansion coefficients and the Lame [equation (C.5)],

Substituting equation (C.31) into equation (30b),

By substituting into equation (C.32) the relationship between the elastic and the Lame constants [equation (C.5)], it is found that

The term in square brackets in the above equation is the specific heat capacity at constant pressure [Relationship between specific heat at constant pressure and specific heat at constant volume], and hence

And finally

Which is the primary relationship of TSA and is valid for homogeneous isotropic bodies undergoing elastic and adiabatic transformations, with the additional assumption that the elastic and thermal properties are constant on temperature [16].

Available Equipment at the IITH.

C-1. IR Camera

Detector Materials : InSb

Spectral Response : 2.5-5.1 μm

Pixel Resolution : 320 x 256

Frame Rate : 380 Hz

Min. Focus Distance : 14 cm

Integration Time : 3 µs to 20 ms

Temperature Range : 5 °C to 300 °C

C-2. Halogen Lamps

Max Capacity : 2.5 kW (230V AC)

Max Capacity : 1.7 kW (115V AC)

Hedler Reflektor : Maxi Norm (7015)

Quantity : 02

C-3. IRX Box (USB)

Interface electronics between camera, software and excitation sources.

Hardware Platform: National Instruments standard components.

Exact hardware synchronization between camera and excitation source.

Digital I/O’s for control functions.

Technical data of Araldite CY 230-1/HY 951

MAT LAB Codes

Contrast Enhancement Techniques

%% Contrast Enhancement Techniques

close all;

clear all;

clc;

%% Step 1: Load Images

I = imread('CFRP_UD_PWH-IMG-002.png'); % file of the image

[X, map] = imread('shadow.tif');

shadow = ind2rgb(X,map); % convert to truecolor

section = rgb2gray(I); % 3D image to 2D image convert

%% Step 2: Enhance Grayscale Images

section_imadjust = imadjust(section);

section_histeq = histeq(section);

section_adapthisteq = adapthisteq(section);

%% Show the Imadjust

J = imrotate(section_imadjust,-180);

figure('position',[100 100 800 600]), imshow(J);

axis on

title('Imadjust');

%% Show the histogram equalization

L = imrotate(section_histeq,-180);

figure('position',[100 100 800 600]), imshow(L);

axis on

title('Histeq');

%% Show the adaptive histogram equalization

N = imrotate(section_adapthisteq,90);

figure('position',[100 100 800 600]), imshow(N);

axis on

title('Adapthisteq');

– THE END –