NDT.net - September 1998,
Table of Contents ECNDT '98
Lock-in Thermography for Nondestructive Evaluation of Aerospace Structures
D. Wu, Th. Zweschper, A. Salerno, and G. Busse*
Institut für Kunststoffprüfung und Kunststoffkunde
Universität Stuttgart, D-70569 Stuttgart-Vaihingen, Germany
|*Corresponding Author Contact:|
G. Busse; firstname.lastname@example.org; http://www.ikp.uni-stuttgart.de
The main goal of nondestructive testing -detection of defects early enough to avoid catastrophic failure- is of particular interest for the inspection of aerospace structures where the failure of a component may cause costs which exceed by many orders of magnitude the cost of the component that failed. Under this aspect all methods for fast and reliable inspection deserve special attention.
It is the aim of this article to present examples that indicate the potential of a method for remote thermal inspection allowing for the detection of defects on m2-sized components within a few minutes. This is important to monitor the change of structural signatures during short inspection intervals.
Thermography is a popular method providing colorful images of areas where local changes of temperature indicate hidden structures and/or functionality of heating or cooling systems. Therefore thermography is a valuable tool for inspection of electrical components, heaters, buildings, and also the human body. As most of these applications rely on the detection of anomalies in the expected temperature pattern, there might be effects of artefacts (e.g. reflections from the environment, air turbulences, inhomogeneous infrared emission coefficient) that give rise to false alarms or, even worse, hide real and critical anomalies. In signal theory it is well known that induced effect modulation provides an efficient means for coding the real signal and thereby to separate it from accidental signals. If this basic concept is applied to thermal inspection, the idea is that heat has to be injected in a modulated or pulsed way and that the resulting change of the thermography image has to be analysed in terms of the parameters of modulation. As heat propagation is slow, one has to be prepared for a response that is slower than with sound waves or even electromagnetic waves (e.g. x-rays). However, the advantage of thermal methods is that they do not need mechanical contact with the inspected component and that one avoids all problems related to physical coupling e.g. of an ultrasonic transducer which prevents rapid raster scan imaging. Also electromagnetic waves do not require mechanical contact, but they may not be applicable in all situations (e.g. the component is too thick for x-ray inspection or too much in motion for interferometric imaging). Modulation thermography methods are not sensitive to vibrations and allow for near-surface inspection with depth information even in thick components. This way they combine the advantages of electromagnetic and elastic waves for remote and nondestructive inspection.
Heat injection during the short time of a pulse (e.g. typically 10 ms for a powerful flash lamp) may cause initially high surface temperatures. Modulated heat deposition /1-4/ distributes energy over a longer period of time and thereby reduces the thermal load of the inspected component to a power density which is typically below the one of sunshine at noon in summer. Such a situation is acceptable for aerospace structures.
The basic idea of lockin thermography is that the temperature modulation induced from the outside on the surface of the inspected component propagates as a "thermal wave". As this wave (describing the space-time-dependence of temperature modulation /5-7/) undergoes reflections at boundaries like all other waves, the temperature modulation at the surface is modified by thermal waves coming back from the inside of the component. A sensitive indicator for such contributions is the phase angle between energy deposition and local thermal response. If the temperature field is monitored during the modulated illumination with a thermography camera, Fourier analysis performed at each pixel provides magnitude and phase of the local response. These two quantities can be used to present the relevant information as another kind of images. The magnitude image is affected by inhomogenities of optical surface absorption, infrared emission and distribution of optical illumination. However, in the phase image each of these effects is eliminated when the evaluation is performed at each pixel. This is of relevance for aerospace structures with typical sizes that make homogeneous optical illumination difficult. If the modulation is performed in a sinusoidal way, this evaluation is particularly simple because averaging procedures reduce the about 1000 initial thermography images to only 4 which are by 90 degrees out of phase giving the phase image according to
where S1 to S4 denote the 4 mentioned images /4/. The sinusoidal modulation can be achieved by a self-learning program performed by the computer that controls the thermography camera together with the lamp power /8/. The range of subsurface detection using thermal waves depends on the frequency of modulation, it increases with the inverse of square root /7/. Signal phase has the advantage that range is by a factor of almost 2 higher than with magnitude /9-12/.
Fig.1 : Lock-in measurement setup
The experimental setup that we used is shown in fig.1 where the lamp (or several of them, each with 1000W) illuminates the component during less than 5 minutes while the thermography system (Agema 900 Lockin /13/) acquires images in the wavelength range 8 - 12 (m at a rate of 15 images/s. Alternatively to the lamps one can also use devices producing air flows of modulated temperature, e.g. an array of computer controled hot air guns.
3. Examples for application of lockin-thermography.
The capability of subsurface feature detection is demonstrated by fig.2 where Teflon foils embedded in various depths of a laminate simulate hidden subsurface defects. The two phase images taken at different modulation frequencies look different since the depth range at the lower frequency is higher thereby showing also deeper defects. Image processing could provide information selectively from certain depths, similarly to tomography methods with electromagnetic or acoustic waves. Investigations were performed on components from various aircrafts (a spacecraft was not available). No surface treatment or application of paint had to be performed prior to inspection with lockin-thermography.
Fig. 2: Embedded Teflon foils in various depths of a CFRP-sample
Lockin thermography imaging of a small airplane wing (Extra 300) has been reported previously /14/ where we showed that the non-uniform illumination affects only the noise in the phase image. So that for this CFRP (carbon-fibre reinforced polymer) wing the structural integrity can be confirmed by such an inspection. Subsurface features were also inspected on an airplane (Grob 115, fig. 3a) made of GFRP (glass fibre reinforced polymer). In the phase image of the nose section (fig.3b) one sees the areas of increased thickness caused by overlapping GFRP-material, while the three bright spots indicate where the cables for the landing light are attached underneath the surface.
Fig. 3:Optical image (a) of an airplane (Grob 115) and phase image (b) of its nose section
We also investigated how well one can use lockin-thermography to inspect larger objects: the fuselage of a large helicopter. The thermography image (fig. 4a) gives an impression of the inspected helicopter area. One sees the effect of non-uniform illumination of the CFRP skin laminate and some indication of subsurface features which are much clearer in the phase image of the same area (fig. 4b). The damage in this fuselage (honeycomb structure partly filled with water, impact damage) appears as spots on the background of known intact structure.
Fig. 4: Thermography image (a) of helicopter fuselage shows nonuniform heating while phase image (b) reveals internal structures and damages
Fig.5: Phase image of an airplane structure with rear surface stringers (horizontal bright lines in the center). Disbonding areas indicated by the interruption of bright lines.
Fig. 5 shows the detection of stringers that lost adhesion with the outer skin: the bright horizontal lines are interrupted where heat transport is reduced.
Besides these examples for monitoring the integrity of structure /14/ we performed measurements aiming at early detection of bolts that are loosening. Phase images obtained on metal plates pressed together by screws fastened at different torques is presented in fig.6 /15/. If the phase profiles across a diameter is performed for each screw, one obtains fig.7 where one sees that signal phase decreases with increasing distance from each screw and that the maximum height depends on torque. Therefrom one can derive a calibration curve showing with which torque an individual screw in an ensemble has been fastened (fig.8). Present investigations correlate this result to the rapid identification of unreliable bolts.
Phase image of metal plates pressed together by screws fastened at different torques.
Fig. 7: Phase angle values plotted along lines in fig.6
Fig. 8: Correlation of phase values with different torque levels
4. Ultrasound lockin-thermography
The detection of defects from the comparison of images showing the inspected component in its initial condition and then at a later inspection requires attention and good knowledge. Hence its reliability may be affected by human factors. Image processing of these images could eliminate this risk and require additional time. Hence it would be highly desirable to distinguish defects from intact areas in a way like dark-field-microscopy which displays only certain kinds of structures.
First attempts to develop such a technique have been performed recently . The basic idea is that defects have a special mechanical behaviour by which they may reveal themselves under certain conditions: their increased hysteresis effect results in selective heating under periodical load, which can be applied by ultrasound the power of which is modulated at the low frequency of the thermal wave.
To apply such a technique one has to inject an elastic wave into the component which is possible only by mechanical coupling since air coupling /17/ has a low efficiency. We pressed a piezoceramic transducer to one edge of the component /18/. Acoustic waves propagate from it into the material where they are attenuated mainly at defects whose heat production is modulated so that they become local sources of low frequency thermal waves.
5. Examples for application of ultrasonic lockin-thermography
The application of this technique is useful in all situations where defects have increased hysteresis effects, e.g. corrosion (internal friction of the material structure) and parts or layers rubbing against each other under oscillation. Fig. 9 shows an aluminum component of an aircraft. In this case the lamp was replaced by an ultrasound source whose amplitude was modulated at the low thermal wave frequency which was again synchronised to the lockin-analysis of the camera. The component under inspection suffered from corrosion on the rear surface. This area appears as a bright spot in the lower right edge of fig. 9. The reason is that the corroded part has a higher loss angle resulting in modulated heat generation when the sample is exposed to ultrasonic excitation. The phase angle indicates the depth where the defect is located, while the magnitude indicates how efficiently the modulated sound is converted into heat (that drives the thermal wave) or how strong the corrosion effect is. The other features of the component (e.g. thickness variation) are suppressed by this principle of selective modulated defect heating. So this technique might be suited for fast and reliable detection of hidden corrosion, delamination, and impact.
|Amplitude (0.12 Hz) Phase (0.12 Hz) |
Fig. 9: Selective heating of corroded area in an alumium airplane component
Fig. 10: Detection of hidden corrosion areas on rear side of the bottom Al-plate.
Another example is presented in fig. 10 where corrosion was not detected in a block of material but in a component where various plates are connected with bolts, a situation that is typical e.g. for fuselage structures. Also in this case corroded areas are revealed with the phase image. It should be noted, that in this arrangement the thermal wave is launched from the defect site itself, hence the images show the waves that are transmitted through the material on top of the defect to the surface.
We presented examples indicating the applicability of modulation thermography methods for the detection of subsurface features and defects in aerospace structures. The inspection requires either a remote source (lamps or warm air) showing the whole structure, or ultrasonic excitation with an attached transducer. In that case the images display only defects and not the intact areas of the structure. As these methods are applicable to large areas within a few minutes, they are suited for fast inspection of structural integrity at depths that are adjustable by modulation frequency. The limit of depth range is presently about 5mm for carbon or glass fibre reinforced materials. For ultrasonic lockin thermography there is still the problem of injecting high power ultrasound into a component without excessive heating of the coupling area. We are still far from the situation where a powerful transducer is clipped to a wing tip and thermography cameras find out which bolts or fasteners need to be replaced.
- Carlomagno G.M., Berardi P.G. (1976): Unsteady thermotopography in non-destructive testing. Proc. 3rd Biannual Information Exchange, Aug. 24-26, St. Louis/USA, 33-39
- Beaudoin J.L., Merienne E., Danjoux R., Egee M. (1985): Numerical system
for infrared scanners and application to the subsurface control of materials by
photothermal radiometry. Infrared Technology and Applications SPIE 590, 287-292
- Kuo P.K., Feng Z.J., Ahmed T., Favro L.D., Thomas R.L., Hartikainen J. (1987): Parallel thermal wave imaging using a vector lock-in video technique. Hess P. and Pelzl J. (Eds.): Photoacoustic and Photothermal Phenomena. Heidelberg. Springer-Verlag 415-418
- Busse G., Wu D., Karpen W. (1992): Thermal wave imaging with phase sensitive modulated thermography. In: J. Appl. Phys. 71, 3962-3965
- Fourier J. (1824): Théorie du mouvement de la chaleur dans les corps solids, 1re Partie, Mémoires de l'Académie des Sciences 185-555
- Angström M.A.J. (1863): New method of determining the thermal conductibility of
bodies. In: Phil. Mag. 25, 130-142
- Rosenwaig A. (1979): Photoacoustic microscopy. American Lab 11, 39-49
- Wu D. (1996): Lockin-Thermographie für die zerstörungsfreie Werkstoffprüfung und Werkstoffcharakterisierung. Dissertation, Universität Stuttgart
- Busse G. (1979): Optoacoustic phase angle measurement for probing a metal. Appl. Phys. Lett 35, 759-760
- Thomas R. L., Pouch J.J., Wong Y. H., Favro L.D., Kuo P. K. and Rosencwaig A. (1980): Subsurface flaw detection in metals by photoacoustic microscopy. J. Appl. Phys. 51 1152-1156
- Lehto A., Jaarinen J., Tiusanen T., Jokinen M., and Luukkala M. (1981): Magnitude and phase in thermal wave imaging. Electr. Lett. 17, 364-365
- Bennett C. A. and Patty R. R. (1982): Thermal wave interferometry: A potential application of the photoacoustic effect. J. Appl. Phys. Lett. 36, 130-132
- BMBF Projekt Nr. 03N8006B0 (1996): Entwicklung thermischer Tomographie mittels
phasenempfindlicher Modulationsthermographie zur zerstörungsfreien Werkstoffprü-
fung von Schichtwerkstoffen
- Wu D., Salerno A., Malter U., Aoki R., Kochendörfer R., Kächele P. K., Woithe K., Pfister K., Busse G. (1996): Inspection of aircraft structural components using lockin-thermography. Proc. QIRT'96, Edizioni ETS, Pisa, 251-256
- Zweschper Th. (1998): Lockin-Thermographie in der Fügetechnik, Fakultät 13, Universität Stuttgart.
- Rantala J., Wu D., Busse G. (1997): NDT of polymer materials using lock-in thermography with water-coupled ultrasonic excitation. NDT&E International, Butterworth-Heinemann Ltd., Oxford, Vol. 31/1, 43-49
- Safaeinili A., Lobkis O. I., Chimenti D. E. (1996): Air-Coupled Ultrasonic Estimation of Viscoelastic Stiffnesses in Plates, IEEE Trans. Ultrason., Ferroelect., and Freq. Contr., Vol. 43, 1171-1180
- Rantala J., Wu D., Salerno A., Busse G. (1996): Lock-in thermography with mechanical loss angle heating at ultrasonic frequencies, 389-393
Copyright © NDT.net,
/DB:Article /SO:ECNDT98 /AU:Wu_D /AU:Salerno_A /AU:Busse_G /CN:DE /CT:IRT /CT:Lockin /CT:Aerospace /ED:1998-09