Back to Basics

[ click here for the Back to Basics Archive ]

Thermography of Composites

by Steven M. Shepard*

 

The potential of infrared thermography as a wide-area, noncontact nondestructive testing (NDT) method has been widely recognized for many years. However, only in the past decade has thermography become popular as a primary NDT method. Prior to that, thermography was primarily used as a qualitative adjunct to traditional NDT technologies, or for preventative and predictive maintenance applications. While it is now used for a variety of NDT applications, the earliest and most enthusiastic acceptance of thermography for NDT has come from the aerospace composites community (Baughman, 1998; Crisman, 1992; Ducar, 1999). Numerous companies, in both commercial and military aerospace, have implemented thermographic testing, often replacing ultrasound or radiography, for both manufacturing and inservice applications for the detection of delaminations, porosity, trapped water and adhesive disbonds. In this article, the principles of thermography are examined as they apply to the NDT of composites, as well as the limitations and potential complications of the technology.

Although thermography is often described as a new or
emerging NDT method, it has a long history.

BASIC PRINCIPLES

In thermography, the surface temperature of a sample is monitored to allow identification of regions where heat flow is modified by subsurface anomalies (Figure 1). To accomplish this, a sample is excited so that heat will flow through it predictably (for example, from one face to the opposite face). The conduction of the heat from the sample surface to the cooler interior is inferred from the changes in surface temperature that are detected, for example, by an infrared camera. The presence of subsurface anomalies in the sample interferes with the expected heat flow, causing transient anomalies in the surface temperature (Vavilov, 1994). For example, heat applied uniformly to the entire surface of a sample flows toward the cooler interior, but may be obstructed by an internal void. As a result, the nearby surface cools at a slower rate than the surface in a distant area free of anomalies. Conversely, an inclusion that conducts heat at a higher rate than the surrounding material will cause the nearby surface to cool more quickly.

While there are many variations on the basic scheme described above, they are based on the same underlying physical principles, and are subject to the same fundamental limitations. Thermographic NDT techniques are active when they are based on analyzing a sample's thermal response to an external excitation, and are passive when based on the temperature of a structure or component under steady-state conditions, as in predictive and preventative maintenance (Maldague, 2002). The differences between the various active approaches to thermographic NDT are basically a matter of what form of excitation is used, and how and where it is applied to the structure.

HISTORICAL PERSPECTIVE

Although thermography is often described as a new or emerging NDT method, it has a long history that predates the development of the infrared camera. In the 1960s, without the benefit of a camera to remotely detect surface temperature changes, the process of thermographic testing was far more laborious and daunting than it is with the dedicated systems that are commercially available today. Wide-area testing typically required that the test object was coated with a temperature-sensitive material, applied by paint or crayon, which could be photographed with a conventional (visible light) camera to provide a permanent record for analysis. Devices such as the evaporograph imaged radiation from the specimen onto a membrane coated with an oil film that generated interference patterns correlated to the object's surface temperature, which could be observed as color variations (McGonnagle, 1961). Cholesteric liquid crystals have been used in some aerospace applications, as they offer the simplicity of paints and tapes, but do not compromise the test surface to the same extent (Williams et al., 1980).

As an alternative to wide-area imaging techniques, single point detectors have also been used to scan the surface of the sample to collect infrared radiation (Levine and Johnson, 1964). In general, the approaches described above had slow response times and limited temperature sensitivity, and offered their best performance in the 2 mm spectral range, thus requiring targets to be heated to significantly elevated temperatures. The combination of these factors and the generally poor performance of thermography compared to ultrasonic and radiographic technology of that period account for its infrequent use.

Figure 1 - Diffusion and detectability: (a) heat generated at the surface propagates into a structure; (b) heat is obstructed by internal anomalies; (c) as the aspect ratio (diameter/depth) of the anomaly decreases, the effect of the anomaly on the surface temperature becomes increasingly difficult to detect, becoming undetectable for temperature differences lower than the noise generated by the camera and background.

Enter the Infrared Camera

The introduction of the infrared camera in the late 1970s significantly altered the prospects for thermographic NDT. The camera eliminated the need for temperature-sensitive pigments or scanning apparatus. Although early cameras relied on optomechanical scanning technology using only one or a few detectors to image a surface, the camera internally handled the mechanical and bookkeeping tasks associated with scanning. The net result was an integrated device that provided a continuously updated infrared image to the operator with unprecedented ease of use. Although the first generation of infrared cameras marked a major step forward in remote temperature sensing, it still left much to be desired in its application to NDT. Early infrared cameras were hampered by low frame rates (approximately 1 Hz), low spatial resolution (approximately 64 x 64 pixels) and poor temperature sensitivity (approximately 1 K [1.8 °F]).

These hardware limitations also limited the types of NDT applications that could be realistically addressed with thermography. For example, highly transient signals, which are likely to occur as a result of anomalies in metals with high thermal diffusivities, were not well suited to the low frame rates of early cameras. In fact, some transient signals associated with anomalies in common engineering materials such as aluminum were likely to occur and vanish in the time interval required to scan a single video frame. Metals posed additional problems in that their low infrared emissivities required the sample surface to be painted black. Even when painted, metals have relatively high thermal effusivities, which meant that for a given thermal excitation, the resulting temperature increase in the metal was relatively low.

The state of infrared camera technology has evolved considerably since the days of the first models. Today, cameras routinely operate at 60 Hz frame rates, and models with rates over 1 kHz are commercially available. While early cameras required liquid nitrogen or large cooling systems, modern cameras are either uncooled, using microbolometer technology, or cooled with miniature stirling engines. Sensitivities as low as 20 mK (0.036 °F) are available in commercial cameras.

In parallel with the progress of infrared camera technology, the dramatic evolution of the personal computer and the development and standardization of high speed data communication between camera and computer have facilitated signal processing algorithms specifically designed for real-time analysis of thermographic NDT signals. These algorithms, which routinely appear in modern NDT systems, significantly extend sensitivity and accuracy beyond results achieved by viewing the unprocessed infrared camera signal (Ringermacher et al., 2001; Shepard et al., 2003; Zweschper et al., 2001). As a result of these developments, modern thermography has evolved to become a precise, quantitative technique that stands alongside established technologies such as ultrasound and radiography.

APPLYING THERMOGRAPHY

There are many factors that determine whether thermography is an appropriate test methodology for a particular application, and, if so, which type of thermography offers the best match for particular application requirements. However, in most cases, the feasibility of a thermographic test can be determined by a few basic considerations.

Thermal Diffusivity

The ratio of thermal conductivity to heat capacity, thermal diffusivity indicates the rate at which heat travels through a material. In high diffusivity materials (for example, copper and aluminum), heat propagates rapidly, to the extent that NDT may require the use of high speed infrared cameras and excitation sources. In low thermal diffusivity materials (for example, glass and rubber), heat propagates slowly and the NDT process may be prohibitively long. Polymer composites generally have intermediate thermal diffusivities that are well suited to NDT with standard frame rate infrared cameras and conventional sources (Table 1).

Thermal Effusivity

The square root of the product of thermal conductivity and heat capacity, thermal effusivity indicates the thermal inertia of a material; that is, its resistance to a temperature increase as heat is applied to it. High thermal effusivity materials, such as metals, do not exhibit large temperature changes in response to excitation, and consequently offer relatively low infrared signal levels that are also susceptible to noise. Polymer composites have relatively low thermal effusivities, and typically provide strong infrared signals that are far above the level of background infrared radiation and instrumentation noise.

Excitation Coupling

Although there are many ways to provide the excitation required for active thermography, it is essential that the excitation is well matched to the application. For example, visible light excitation is effective for optically opaque samples, but not for translucent or reflective samples (unless they are coated with an opaque paint or membrane). Similarly, induction or microwave excitation is appropriate for samples having some metallic content, but not for samples that are dielectric (for example, glass fiber reinforced polymeric composites).

Limits of Detectability

Although the changes in the sample surface temperature are measured by an infrared camera, in most cases, the primary limits of detectability of subsurface features or anomalies are not due to the optics or pixel density of the infrared detector. Rather, the most severe constraints on thermography are usually due to the fact that signals occur as a result of a diffusion process — specifically, the diffusion of heat through the sample (Figure 2). The net effect of thermal diffusion on the NDT process is that the minimum detectable feature size increases with the depth of the anomaly or feature, and the time required to detect a feature is proportional to the square of its depth. Most importantly, there is a limit to the minimum detectable anomaly at any depth, which is expressed in terms of the anomaly aspect ratio (the ratio of diameter to depth): anomaly diameter/anomaly depth > approximately 2 (Maldague, 2002). The exact limit imposed by the aspect ratio will vary depending on the excitation technique, the infrared camera performance, the sample material and the nature of the anomaly (for example, delamination or inclusion). It is important to realize that this is a best-case indicator that does not account for camera performance or the amount of energy used to excite the sample. However, when considering thermography, it is important to determine that the application requirements, even under the most favorable conditions, are realistic for these methods. For example, detecting a 1 mm (0.04 in.) diameter void at a depth of 10 mm (0.4 in.), where the aspect ratio is 0.1, is highly unrealistic for thermographic NDT techniques, while detecting a 10 mm (0.4 in.) diameter void at 1 mm (0.04 in.) depth (aspect ratio = 10) is quite realistic.

Figure 2 - Thermal excitation may be applied to the surface: (a) using light or hot air; (b) using internal electromagnetic induction or microwave heating; (c) at anomaly site using mechanical or sonic energy.

Table 1 - Thermophysical properties of selected materials

Emissivity

Regardless of the excitation technique employed or the sample material involved, in most modern thermographic practice, changes in the surface temperature are detected by an infrared camera. This requires that the surface of the sample is an effective emitter of infrared energy. Emissivity is the ratio of the infrared energy radiated by an object at a given temperature to the energy emitted by a perfect radiator (emissivity = 1), or blackbody, at the same temperature, and thus, is an excellent measure of how effectively an object emits infrared energy. Emissivity is a property of the sample surface only, so that low emissivity can usually be remedied by preparing the surface with a high emissivity coating (for example, washable black paint). Thus, bare metals, which are notoriously poor emitters (emissivity << 1), may require surface preparation, while most polymer composites used in aerospace applications have sufficient emissivity to allow testing without surface preparation. However, occasionally one will encounter materials that are infrared translucent (such that the camera "sees through" the sample), which are difficult to test unless an opaque coating is applied.

EXCITATION METHODS

The range of techniques used to excite a sample for thermography include radiative heating (for example, light, infrared, microwave, electromagnetic induction), mechanical stimulation (sonic/ultrasonic, cyclic stress, convection, direct contact with hot/cold source), as well as chemical (exothermic reaction of a binary adhesive) and electrical (joule heating). The excitation may be applied to the surface of the sample (for example, light or direct contact excitation) or internally (induction, sonic or chemical heating). Thermography may be performed as a single-sided or dual-sided access technique, depending on whether the excitation source and camera are on the same or opposite sides of the sample. As a practical matter, single-sided testing is often preferred, since both sides of complex structures may not be accessible.

Optical Methods

Currently, optical excitation techniques are the most widely used form of thermography in both the manufacturing and maintenance of composite aerospace structures. Using visible light to heat the surface, optical excitation offers noncontact, wide-area performance that is comparable to the performance offered by ultrasonic testing for many applications (Cargill et al., 2006). Optical excitation is a particularly attractive alternative for testing of structures, where scanning a large area is involved. In most cases, optical excitation systems do not require critical alignment and positioning, so that a large scale thermographic system is less complex (and less expensive) to implement than, say, an ultrasonic system, where precise multiaxis alignment and positioning must be maintained.

Typically, optical excitation for thermography is performed using either xenon flash lamps or halogen heat lamps. The excitation is delivered as either a pulse, a step, a modulated (on/off) sequence, or along a continuous line or area that is scanned over the sample surface (Maldague, 1993). In comparing these techniques, it is easy to lose sight of the fact that although they represent different approaches to signal processing, the fundamental interaction occurring in the presence of an internal obstruction of incident heat flow is identical for all of them.

The choice of optical excitation technique is usually dependent on the nature of the anomaly to be detected (size and material of the anticipated discontinuity), the scale of the application (test time and area to be covered), and whether automation and/or quantification will be required. For example, very large scale applications, where severe anomalies are likely, but not requiring quantification or automation, might be better served with a continuous or modulated source. However, production applications, where speed, quantification and automation are required, are often more compatible with flash sources. Today, flash thermography is the most widely implemented form of thermographic NDT. For most applications, it is capable of testing approximately 0.1 m² (1 ft²) in less than 10 s, and can be fully automated.

Sonic Excitation

While optical methods apply energy to the sample surface to initiate internal heat flow, sonic excitation techniques use an entirely different principle. The sample is observed with an infrared camera as it is mechanically stressed, which generates a negligible temperature increase for an ideal, anomaly-free sample. However, the presence of cracks or interfaces that rub or slap against each other as mechanical stress is applied may result in a temperature rise at the surface of the sample (Favro et al., 2000; Henneke et al., 1979; Mignogna et al., 1981). Thus, sonic thermography is selective in that, in principle, a signal occurs at the surface only if an anomaly is present.

In vibrothermography (also known as sonic thermography), stress is applied to the sample via an ultrasonic horn, typically at low frequencies (15 to 30 kHz) and at high power (1 to 2 kW). While it is possible to detect anomalies such as delamination or debonding in composites using vibrothermography, precautions must be taken to avoid damaging the sample surface with the horn, and also to find an appropriate insertion point for the sonic energy that will not create blind spots where anomalies are undetectable (Shepard et al., 2004).

Electromagnetic Heating

Techniques such as induction or microwave heating can be applied to materials with sufficient metallic content (for example, a honeycomb core supporting a glass fiber laminate). Electromagnetic excitation can be used to heat the interior rather than the surface of a sample (the penetration depth of the heating being inversely proportional to the frequency of the electromagnetic radiation). While electromagnetic techniques can penetrate deep into a sample, the surface temperature signal associated with a subsurface anomaly is still subject to the limitations imposed by diffusion (Lehtiniemi and Hartikainen, 1994; McCullough, 2004; Osiander et al., 1995).

THERMOGRAPHY AND COMPOSITES

At about the same time infrared cameras became commercially available, composite materials were introduced as an alternative to metals, most notably in the aerospace industry. Structures such as control surfaces and engine cowlings, constructed with carbon-epoxy composites, offered strength comparable to metals with the added advantage of significant weight reduction. From an NDT perspective, however, the metal versus composite comparison raised additional issues. NDT of predominantly aluminum alloy aircraft had been performed using visual techniques. Armed with a flashlight and mirror, an experienced inspector could identify pillowing associated with subsurface corrosion, dents due to impact, or nicotine residue due to disbonding of lap or butt seams (in the days when smoking was common on aircraft). With composite construction, where damage was likely to be undetectable at the surface, visual testing techniques were of little value, so alternatives capable of wide-area testing were sought.

Unlike metals, in many respects composite materials were ideally suited for thermographic testing with the newly introduced infrared cameras (Milne and Reynolds, 1985). With relatively low thermal diffusivities (compared with aluminum), the time scale at which transients associated with anomalies occurred was on the order of seconds, and well matched to camera frame rates. The lower thermal effusivities of composites resulted in greater temperature increases in response to excitation, and the emissivity of most composites used in aerospace manufacture was sufficient to allow testing without painting the surfaces black (Plotnikov and Winfree, 1999).

Anisotropy

Composite laminates are often thermally anisotropic, since thermal conductivity is usually on the order of 10 times higher in the direction along the orientation of the fibers than in the direction perpendicular to them. This raises an apparent contradiction for thermographic schemes, where heat is generated at the surface of a sample and expected to flow to the opposite surface — that is, in the direction perpendicular to the fiber orientation - as it would seem that heat would be most likely to flow laterally, through the fibers, rather than toward the opposite face. This is the situation, indeed, but the contradiction is resolved by the fact that heat is applied to an area of the surface that is considerably larger than the thickness of the sample.

For a point heat source applied to the surface of a sample, the effect of anisotropy would be immediately obvious, as the applied heat would flow along the fibers toward cooler regions, and little of the applied energy would reach the back surface (Figure 3). However, when a larger area is heated, only the region near the edges of the heated area is exposed to the cooler surrounding material, so that the flux out of the heated zone occurs near its periphery. As heat continues to escape the heated zone, heat flow along the fiber direction dominates. However, until that time, the region directly below the heat source behaves as though heat conduction occurs primarily in the direction perpendicular to the surface, allowing interrogation of that material using thermography. Thus, thermal diffusivity in the through-thickness direction can be isolated from the lateral diffusivity components and determined by measuring the time required for a pulse of heat applied to the front surface to reach the back surface of a sample.

While wide-area heating allows thermography to essentially ignore the lateral effects of anisotropy, point or line source heating may be used to detect or even measure anisotropy of the in-plane thermal diffusivity. In fact, many schemes have been developed to measure in-plane diffusivity by observing the lateral flow of heat from a structured heat source at the sample surface (Graham et al., 1998; Ouyang et al., 1998; Sun et al., 1999).

Figure 3 - Anisotropy of thermal conductivity: (a) for a point heat source at a structure's upper surface, heat travels preferentially along the lamina plane; (b) for a plane heat source, conduction along the lamina plane is apparent at the edges of the heated area, but primary flux is perpendicular to the lamina plane, toward the lower surface.

Figure 4 - Impact damage on a graphite epoxy laminate rudder: (a) immediately (17 ms) after flash heating, the temperature on the structure's surface appears to be uniform; (b) an indication of a subsurface anomaly, oriented along the lamina plane, appears later in processed sequence (images courtesy of Thermal Wave Imaging, Inc.).

Figure 5 - Examples of trapped water revealed by thermographic testing of honeycomb cells on composite helicopter rotor blades.

Figure 6 - Delamination in graphite epoxy component (image courtesy of Thermal Wave Imaging, Inc.).

Composite NDT Applications

Although thermography is currently used as a primary NDT technique for a variety of materials and applications, the testing of polymeric composites is the most popular application of thermographic NDT. Flash excitation is most widely used in commercial installations, as it is faster and more compatible with automation and production environments than many of the alternatives. Applications in manufacturing include detection of foreign object debris (such as pieces of polymer film left on prepreg layers), detection of anomalous resin distribution, measurement of porosity, confirmation of ply orientation, and detection of voids. In maintenance, tests include detection of impact damage (Figure 4), trapped water (Figure 5), delamination (Figure 6) and the validation of patch repairs. Further, a multidisciplinary thermographic application was devised by Williams and Nagem (1983) who combined liquid crystal thermography, materials characterization and fracture mechanics to develop a kit for the NDT of composite pleasure boats.

THE FUTURE

Advancements in signal processing and infrared camera technology have made it possible to extend the boundaries of thermography beyond anomaly detection to include materials characterization. In the past, the practice of thermography was based almost entirely on visual assessment of the infrared camera image sequence. Evaluation of results depended on the presence of contrast between an anomaly and an anomaly-free background. Today, each pixel in the camera data stream is treated as an independent entity that can be evaluated without reliance on visual assessment or prior knowledge of the physical characteristics of the sample. Features that are undetectable using conventional image processing techniques are routinely detected using single pixel analysis tools that operate on the camera data in real time.

As the trend toward using composites in critical structural applications continues, the challenge for thermography will be to confront the fundamental limitations imposed by diffusion that make it impractical for testing of massive multilayer structures. A brute force approach (for example, more excitation energy and more sensitive cameras) is not likely to solve the problem and, in fact, may exacerbate the complications. Researchers are actively investigating alternative excitation schemes that will hopefully result in the next generation of thermographic NDT.

REFERENCES

Baughman, S.R., "Applications for Thermal NDT on Advanced Composites in Aerospace Structures," Proceedings of SPIE Thermosense XX, J.R. Snell and R.N. Wurzbach, eds., Vol. 3361, April 1998, pp. 311-319.

Cargill, J.S., J.A. Pecina, L. Gintert and D. Skelton, "Evaluation of NDT Technologies for Organic Matrix Composite Testing," Materials Evaluation, Vol. 64, 2006, pp. 857-864.

Crisman, E.M., personal communication, 1992.

Ducar, R.J., "Pulsed Thermographic Inspection and Application in Commercial Aircraft Repair," Proceedings of SPIE Thermosense XXI, D.H. LeMieux and J.R. Snell, eds., Vol. 3700, April 1999, pp. 77-83.

Favro, L.D., X. Han, Z. Ouyang, G. Sun, H. Sui and R.L. Thomas, "Infrared Imaging of Defects Heated by a Sonic Pulse," Review of Scientific Instruments, Vol. 71, No. 6, 2000, pp. 2418-2421.

Graham, S., D.L. McDowell and R.B. Dinwiddie, "In-plane Thermal Diffusivity Measurements of Orthotropic Materials," Thermal Conductivity: Thermal Expansion 12: Joint Conferences, P.S. Gaal et al., eds., Lancaster, Pennsylvania, Technomic Publishing, 1998, pp. 241-252.

Henneke, E.G., K.L. Reifsnider and W.W. Stinchcomb, "Thermography, an NDI Method for Damage Detection," Journal of Metals, Vol. 31, No. 9, 1979, pp. 11-15.

Lehtiniemi, R. and J. Hartikainen, "An Application of Induction Heating for Fast Thermal Nondestructive Evaluation," Review of Scientific Instruments, Vol. 65, No. 6, 1994, pp. 2099-2101.

Levine, M.B. and B.A. Johnson, "Dynamic Infrared Inspection Techniques," Proceedings of the 4th International Conference on NDT, British National Committee for NDT, Butterworths, London, 1964, pp. 308-314.

Maldague, X.P.V., Nondestructive Evaluation of Materials by Infrared Thermography, London, Springer-Verlag, 1993.

Maldague, X.P.V., "Introduction to NDT by Active Infrared Thermography," Materials Evaluation, Vol. 6, 2002, pp. 1060-1073.

McCullough, R.W., "Transient Thermographic Technique for NDI of Aerospace Composites," Proceedings of SPIE Thermosense XXVI, D.D. Burleigh et al., eds., Vol. 5405, 2004, pp. 390-402.

McGonnagle, W.J., Nondestructive Testing, New York, McGraw-Hill, 1961.

Mignogna, R.B., R.E. Green, J.E. Duke, E.G. Henneke and K.L. Reifsnider, "Thermographic Investigation of High Power Ultrasonic Heating in Materials," Journal of Ultrasonics, Vol. 19, No. 7, 1981, pp. 159-163.

Milne, J.M. and W.N. Reynolds, "Nondestructive Evaluation of Composites and Other Materials by Thermal Pulse Video Thermography," Proceedings of SPIE Thermosense VII, Vol. 520, A.G. Kantsios, ed., Bellingham, Washington, 1985, p. 119.

Osiander, R., J.W.M. Spicer and J.C. Murphy, "Thermal Nondestructive Evaluation Using Microwave Sources," Materials Evaluation, Vol. 53, 1995, pp. 942-948.

Ouyang, Z., L. Wang, L.D. Favro and R.L. Thomas, "Novel Measurement of Anisotropic Thermal Diffusivity," Review of Progress in Quantitative Nondestructive Evaluation, Vol. 17A, D.O. Thompson and D.E. Chimenti, eds., Plenum Press, New York, 1998, pp. 453-456.

Plotnikov, Y. and W. Winfree, "Visualization of Subsurface Defects in Composites Using a Focal Plane Array Infrared Camera," Proceedings of SPIE Thermosense XXI, Vol. 3700, D. LeMieux and J.R. Snell, eds., 1999, pp. 26-31.

Ringermacher, H.I. and D.R. Howard, "Synthetic Thermal Time-of-flight (STTOF) Depth Imaging," Review of Progress in Quantitative Nondestructive Evaluation, Vol. 20A, D.O. Thompson and D.E. Chimenti, eds., Melville, New York, AIP, 2001, pp. 487-491.

Shepard, S.M., J.R. Lhota, B.A. Rubadeux, D. Wang and T. Ahmed, "Reconstruction and Enhancement of Active Thermographic Image Sequences," Optical Engineering, Vol. 42, No. 5, 2003, pp. 1337-1342.

Shepard, S.M., J.R. Lhota and T. Ahmed, "Experimental Considerations in Vibrothermography," Proceedings of SPIE Thermosense XXVI, D. Burleigh et al., eds., Vol. 5405, 2004, pp. 332-335.

Sun, J.G., C. Deemer and W.A. Ellingson, "Thermal Imaging Measurement and Correlation of Thermal Diffusivity in Continuous Fiber Ceramic Composites," Thermal Conductivity, P.S. Gaal et al., eds., Lancaster, Pennsylvania, Technomic Publishing, 1999, pp. 616-622.

Vavilov, V., "Infrared Techniques for Materials Analysis and Nondestructive Testing," Infrared Methodology and Technology, Monograph Series International Advances in NDT, X. Maldague, ed., Gordon and Breach, 1994, pp. 230-309.

Williams, J.H., Jr. and R.J. Nagem, "A Liquid Crystals Kit for Structural Integrity Assessment of Fiberglass Watercraft," Materials Evaluation, Vol. 41, No. 2, 1983, pp. 202-210.

Williams, J.H., Jr., S. Hossein Mansouri and S.S. Lee, "Thermal Non-destructive Testing of Fiberglass Laminates Using Liquid Crystals," British Journal of NDT, Vol. 22, No. 3, 1980, pp. 113-118.

Zweschper, T., A. Dillenz and G. Busse, "NDE of Adhesive Joints and Riveted Structures with Lock-in Thermography Methods," Proceedings of SPIE Thermosense XXIII, A. Rozlosnik and R.B. Dinwiddie, eds., Vol. 4360, 2001, pp. 567-573.

 

* Thermal Wave Imaging, 845 Livernois St., Ferndale, MI 48220; (248) 414-3730; fax (248) 414-3764; e-mail sshepard@thermalwave.com.

 

Copyright © 2007 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Back to Materials Evaluation ]