Laser NDT techniques have proven to be a valuable and efficient tool for solving complex mechanical problems.

These speckles can be unwanted (for example, if a clear image of the object is required), but they also carry important information that can be used for measuring purposes (Jones and Wykes, 1989).

In principle, the speckle image is akin to a fingerprint of the object surface. It includes information about the position and contour of the surface in encoded form. It is difficult to derive this information from the speckle image directly, but we can use it to measure displacements and deformations.

Laser Speckle Correlation
If the object in Figure 1 is moved laterally to the camera position, the speckles will move in the image accordingly. Therefore, if we take an image in the nondeformed state of the object and one in the deformed state and manage to find the new position of the speckle pattern, we know the lateral displacement of the object. This can be automatically done by an image correlation calculation between the two images (Sutton et al., 2000).

In practical applications, however, the object will normally not only move, but also distort, causing the speckle pattern to change as well. This will result in speckle patterns according to Figure 2. The magnification of the complete speckle pattern of the object (top) shows the detailed speckle distribution in the nondeformed state (bottom left). After applying the load, the correlation algorithm will detect the new position of the speckle pattern, which is dislocated and distorted (bottom right). In this way, we receive the information both about the lateral displacement and the distortion respective of the strain on the surface. Since this procedure is done automatically for the complete image, we receive a complete map of displacements and strains with this technique.

Of course, knowing the exact camera magnification is necessary to receive accurate information about the object deformation. Therefore, prior to deformation measurement, the magnification has to be measured, typically by using calibration boards with precision markings. The accuracy of this technique is basically determined by the number of pixels of the camera and the size of the object being measured. In practice, we can assume an accuracy of around 1 to 10 µm (3.9 ¥ 10-5 to 3.9 ¥ 10-4 in.) for object sizes from 0.1 m to 1 m (3.9 to 39.4 in.) in diameter. In terms of strains, strain rates of 100 µm/m (1.2 x 10-3 in./ft) can be resolved.

Figure 3 shows an example of a dog bone test of a small specimen of aluminum alloy (Herbst et al., 2003). In Figure 3a, the original speckle image in the nondeformed state is seen. The specimen was pulled in the vertical direction and a series of images was recorded up to fracture. The images were analyzed, showing the displacements and strains of each surface point at each load step. Figure 3b shows the strain distribution at maximum load just before fracture. As the series of images was recorded, the evolution of the strains in comparison to the applied force can also be deducted for every surface point. For example, Figures 3c and 3d show the evolution of the longitudinal and transverse strains at point A, the fracture point.

In comparison to classical strain gage techniques, the advantage of such a technique is obvious. Full-field digital laser correlation can deliver a complete set of information at the same time, including

• amplitude of longitudinal and transverse strain at every load level (this also allows the determination of the poisson ratio)
• average strain and peak strain
• point of maximum strain/point of fracture (this is important in fracture tests, where the position of the later fracture is not known in advance).

It should be mentioned that digital laser correlation leads to significant amounts of data, which typically cannot be processed online. Therefore, the results of the measurement are typically analyzed in a separate process. Also, accuracy of digital laser correlation cannot yet reach the accuracy of classical strain gage measurements. However, as camera resolution and computer processor power increase, both limitations should be overcome in the future.

Laser Speckle Interferometry Laser speckle interferometry (whether electronic speckle pattern interferometry [ESPI] or digital speckle pattern interferometry [DSPI]) also uses the laser light that is scattered back from a surface. However, in comparison with laser correlation, the speckle pattern is superimposed with a second light wave, producing an interference pattern (Yang and Ettemeyer, 2003).

In the optical setup in Figure 4a, the laser beam is split into different directions, forming several laser illumination beams and a reference beam. Combining different pairs of such laser beams and recording an image at both the nondeformed and deformed state, this setup can measure the out-of-plane component of the deformation as well as its in-plane component.

For example, the bending of a circumferentially fixed plate will produce an out-of-plane deformation according to Figure 4b. The image in the top right shows the deformation as lines of constant amplitude, which are automatically demodulated to the deformation image at bottom right. The amplitude difference between a pair of fringes is a fraction of the laser wavelength. Consequently, the accuracy of ESPI measurement can reach 30 to 50 nm in practical applications: little more than a millionth of an inch!

Using the different laser illumination directions of the ESPI setup, this system can obtain full three-dimensional deformation information. The example in Figure 5 shows a steel sheet with a central hole, which is pulled by the testing machine. The ESPI system with the camera in the center is looking at the test piece and the four illumination directions allow the measurement of the two in-plane components in the x and y direction and the out-of-plane component in the z direction. The image on the monitor in the background shows the strain distribution on the surface of the specimen. The colors indicate the amplitude of the longitudinal strain, showing the strain concentration around the hole and under 45 degrees to the tensile direction.

The advantage of this technique is the high sensitivity in combination with the complete 3D measuring information on the whole measuring surface. In comparison with digital laser correlation, ESPI is typically 10 times more sensitive. This enables quantitative measurements at small strain levels (above 10 microstrains). Practical applications can range from small objects of less than 1 mm (0.04 in.) in diameter up to larger objects several hundred millimeters (a couple feet or so) in length.

Figure 6 shows the results of a tensile test of a composite structure made from carbon fiber reinforced plastics (Schubach et al., 2001). The longitudinal strain distribution is measured on the whole surface. Additionally, the displacement fields show local disturbances well before fracture of the sample (indicated by the arrows).

Three-dimensional ESPI has successfully been used in component tests, fatigue investigations, analysis of thermal expansion, deformation, strain/stress, residual strain/stress and vibration, and in nondestructive testing (NDT) applications. The technique can also be used for measuring the shape of the test object, opening up the possibility of measuring 3D displacements and strains on curved surfaces, such as in component testing (Ettemeyer et al., 1997; Pedrini and Tiziani, 1994; Wegner et al., 2001).

An example for a dynamic application of ESPI is a car disk brake, vibrating at 1.890 Hz (Figure 7; Krupka et al., 2003). The ESPI system is triggered according to the vibration frequency and records the vibration amplitude at every point of the disk. The advantage of this application is the noncontact and full-field nature of the measurement, which avoids the application of many accelerometers. Also, the application is very fast and shows the complete vibration mode, no matter its complexity.

When compared, ESPI and laser speckle correlation offer similar information (Ettemeyer, 2004). Due to its higher resolution, ESPI can resolve smaller strain levels, but on the other hand ESPI is also more sensitive to environmental disturbances, such as unwanted vibrations. Also the application of ESPI is limited to areas up to 90 000 mm2 (139.5 in.2), while laser speckle correlation shows basically no limitation in size.

Shearography
Digital shearography has been developed especially for nondestructive testing purposes (Yang et al., 1995). Again, the object is illuminated by a laser wave and recorded by a camera (Figure 8a). However, before the image reaches the camera, it is doubled, laterally sheared and superposed on the camera. Therefore, we achieve a double image of the object. The outcome of this effect is that shearography measures (as with ESPI) the out-of-plane deformation of the object. However, because surface points with a lateral distance of x? are superimposed, the relative deformation between neighboring object points is obtained. This is, mathematically speaking, the deformation gradient.

In comparison, the deformation of the same plate as in Figure 4b now shows the deformation gradient (Figure 8b). This effect has an important advantage: while still measuring the deformation with the same sensitivity (30 to 50 nm), shearography will not record rigid body displacements. Consequently, its sensitivity towards environmental influences such as unwanted vibrations is significantly reduced, making shearography a practical tool for the work floor. At the same time, the deformation gradient is a much clearer signal for local inhomogenities (those caused by delaminations, voids and cracks in components rather than the deformation itself). Consequently, shearography is mostly used for NDT applications.

For example, if a composite structure is loaded by heat or mechanical stress, the surface layers will deform in the presence of a local disbond (Figure 9a). As shearography shows the gradient of the displacement in the shear direction, the result of a typical delamination will be a butterfly pattern. In Figure 9b, a composite panel with delaminations of different sizes was tested by shearography.

In order to detect discontinuities, an appropriate loading technique is required. Due to the high sensitivity of shearography, the applied load is very small. In the past, the following techniques have proved to be efficient for shearographic testing:

• heating (a difference of a few kelvin [several degrees fahrenheit] can be enough to cause a deformation of a debonded surface)
• vacuum (either in stationary vacuum chambers or with portable vacuum hoods, which stick to the testing surface and cause a little mechanical stress on the surface)
• vibration (typically with piezo shakers, which excite the structure in the test area - the frequency is sweeped, causing significant surface vibrations, if a resonance is hit)
• mechnical stress.

Shearography can be used in automatic testing systems as well as in portable systems for field testing. For example, a fully automatic helicopter rotor blade testing system has been developed by Erne et al. (1999). The rotor blades are positioned in a vacuum chamber and loaded with a vacuum of under 100 mbar while two shearography cameras scan the complete blade. The complete test cycle including loading of the blade takes less than 15 min and all delaminations larger than 15 mm (0.6 in.) in diameter are recorded.

For field testing, a portable shearographic system can be used on a mobile tripod. Loading can be applied by heating with a halogen lamp (Walz et al., 2003) or by mechanical stress. Such tests require minimum preparation and still deliver very precise information about the state of the component.

The main advantage of shearography is its high test efficiency. The high test speed (up to 1 m2/min [10.8 ft2/min]) and a high reliability of finding all discontinuities within the tested area (full-field technique) make it an interesting testing technique. It is best suited to detect delaminations of composite structures close to the surface. Nowadays shearography is mainly used in aerospace applications, but it is also used on marine structures and other transportation industries are starting to investigate the application of this technique.

Conclusion
Laser NDT techniques have proven to be a valuable and efficient tool for solving complex mechanical problems. They are also useful for standard applications, such as material testing, fracture investigations, strain/stress analysis, vibration and acoustics, and NDT of composite components in the aerospace, marine and automotive industries. The full-field information provided by these techniques gives maximum safety and test speed. These data fit perfectly with modern 3D simulation tools, which produce exactly the same results.

The simplicity of operation in combination with the development of interfaces to standard applications will further increase the number of applications of these techniques in the future.

References
Erne, O., T. Walz and A. Ettemeyer, "Composite Structural Integrity NDI with Automatic Shearography Measurements," ASNT Fall Conference, Phoenix, Arizona, October 1999.

Ettemeyer, A., "Material and Component Validation by Speckle Interferometry and Correlation Methods," 16th WCNDT, Montreal, Canada, 2004.

Ettemeyer, A., Z. Wang and T. Walz, "Applications of 3D Speckle Interferometry to Material and Component Testing," Proceedings of SPIE, Vol. 3098, 1997, pp. 188-194.

Herbst, C., K. Splitthof and A. Ettemeyer, "New Features in Digital Image Correlation Techniques," Fourth French Colloquium, CMOI Conference, "Méthodes et techniques optiques pour l'industrie," Belfort, France, November 2003.

Jones, R. and C. Wykes, Holographic and Speckle Interferometry, Cambridge, Cambridge University Press, 1989.

Krupka, R., T. Walz, and A. Ettemeyer, "Fast and Full-field Measurement of Brake Squeal Using Pulsed ESPI Technique," Optical Engineering, Vol. 42, May 2003, pp. 1354-1359.

Pedrini, G. and H.J. Tiziani, "Double Pulse Electronic Speckle Interferometry for Vibration Analysis," Applied Optics, Vol. 33, 1994, pp. 7857-7863.

Schubach, H.R., A. Ettemeyer, R. Krupka and R. Wegner, "Laser-shearografie zur Materialcharakterisierung und Ganzflächigen, Zerstörungsfreien Prüfung von Composite-bauteilen," 6. Kolloquium Qualitätssicherung durch Werkstoffprüfung, Zwickau, Germany, November 2001.

Sutton, M.A., S.R. McNeil, J.D. Helm and Y.J. Chao, "Advances in Two-dimensional and Three-dimensional Computer Vision for Shape and Deformation Measurements," Photomechanics, P.K. Rastogi, ed., Topics in Applied Physics, Vol. 77, New York, Springer, 2000, pp. 323-372.

Walz, T., R. Krupka and A. Ettemeyer, "Industrial Applications of Shearography for Inspection of Aircraft Components," NDT-CE, Berlin, Germany, September 2003. Wegner, R., T. Siebert and A. Ettemeyer, "Combine Simulation and Experiment in Automotive Testing with ESPI Measurement," Danubia Adria Symposium, Steyr, Austria, September 2001.

Yang, L. and A. Ettemeyer, "Strain Measurement by 3D-electronic Speckle Pattern Interferometry: Potentials, Limitations and Applications," Optical Engineering, Vol. 42, 2003, pp. 1257-1266.

Yang, L.X., W. Steinchen, M. Schuth and G. Kupfer, "Precision Measurement and Nondestructive Testing by Means of Digital Phase Shifting Speckle Pattern and Speckle Pattern Shearing Interferometry," Measurement, Vol. 16, 1995, pp. 149-160.

 

 

* Munich University of Applied Sciences, Lothstrasse 34, 80355 Munich, Germany; 49 89 1265 1131; fax 49 89 1265 1480; e-mail andreas.ettemeyer@fhm.edu.

 

Copyright © 2006 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]