NONDESTRUCTIVE INSPECTION OF A COMPOSITE MATERIAL SAMPLE USING A LASER ULTRASONICS SYSTEM WITH A BEAM HOMOGENIZER

Transcription

1 NONDESTRUCTIVE INSPECTION OF A COMPOSITE MATERIAL SAMPLE USING A LASER ULTRASONICS SYSTEM WITH A BEAM HOMOGENIZER J. M. S. Sakamoto 1, 4, A. Baba 2, B. R. Tittmann 3, J. Mulry 3, M. Kropf, 3 and G. M. Pacheco 1 1 Department of Microwave and Optoelectronics, Aeronautics Institute of Technology, São José dos Campos, SP, Brazil 2 Energy and Environmental Systems Laboratory, Hitachi Ltd., Hitachi, Ibaraki, Japan 3 Department of Engineering Sciences and Mechanics, Pennsylvania State University, University Park, PA, USA 4 Division of Photonics, Institute for Advanced Studies, São José dos Campos, SP, Brazil ABSTRACT. A laser ultrasonics system was used to detect impact damage in samples of composite material used in the Brazilian aeronautics industry. The ultrasonic generation was accomplished with a high power Nd:YAG laser at 1064 nm and the detector was a Mach-Zehnder interferometer. The Nd:YAG laser was configured so as not to cause damage to the sample and a beam homogenizer was used to distribute the laser energy homogeneously over the optical spot cross section. Keywords: Laser Ultrasonics, Beam Homogenizer, Composite Material PACS: hc. INTRODUCTION The use and development of carbon fiber composite materials in the Brazilian aeronautics industry in Brazil is bringing a demand for nondestructive inspection methods. Even for well established materials as aluminum there is a need for noncontact inspection for example to inspect weld during the welding process. An all optical method is interesting because it can accomplish the inspection without contact and it doesn t require a coupling gel. Also, it can be nondestructive and can be scanned using suitable optics. In this work it is shown a laser ultrasonics system with a beam splitter and a beam homogenizer to reduce the fluency (surface energy density, J/cm 2 ) of the laser beam in order to inspect and generate images of composite samples without damaging it. The sample inspected was developed in Brazil and it has carbon fiber tissue layers with thickness of approximately 210 µm and the matrix is made of epoxy. There are ten layers Review of Progress in Quantitative Nondestructive Evaluation, Volume 30 AIP Conf. Proc. 1335, (2011); doi: / American Institute of Physics /$

2 (a) (b) FIGURE 1. Composite material sample used in Brazilian aeronautics industry. (a) Front view; (b) Back view. and the fibers orientation is [±45 o, (0 o, 90 o ), ±45 o, (0 o, 90 o ), ±45 o ]s. The center of the sample was damaged by the impact of a dropping mass of 1.5 kg with energy of 7 J. This caused an internal defect that will be imaged using the laser ultrasonics system. Figure 1 shows the front and the back of the composite sample. EXPERIMENTAL SETUP All the experiments were accomplished at the Pennsylvania State University in the Acoustic-optics Laboratory. A Nd:YAG laser operating at 1064 nm wavelength was used to generate ultrasound waves in the composite material. Initially we set the Q-switch delay of the laser to it provides low values of laser energy. The laser spot was unfocused to decrease the fluency. The laser beam was directed to the sample and the single shot mode was used to impinge it. Single shots were given to six different positions on the sample. The fluency used to each shot was 0.34 J/cm 2, 0.45 J/cm 2, 0.56 J/cm 2, 0.67 J/cm 2, 0.78 J/cm 2, and 0.89 J/cm 2. Even for the minimum value of fluency, the laser caused a surface damage on the sample. The damage threshold according to the reference [1] is approximately 0.35 J/cm 2. Figure 2 shows the photographs of the regions impinged by the laser, using an optical microscope. In order to avoid damaging the sample, the fluency was decreased including a beam splitter on the setup. A beam homogenizer was also used to improve the distribution of the energy over the laser spot cross section. The homogenizer used was made of a dielectric (a) (b) (c) (d) (e) (f) FIGURE 2. Regions impinged by the laser. Fluency: (a) 0.34 J/cm 2 ; (b) 0.45 J/cm 2 ; (c) 0.56 J/cm 2 ; (d) 0.67 J/cm 2 ; (e) 0.78 J/cm 2 ; (f) 0.89 J/cm

3 FIGURE 3. Experimental setup. material (BK7), had a hexagonal cross section with 2 mm of entrance and exit aperture and 75 mm length. The setup with the beam splitter and the homogenizer is shown in Fig. 3. The light provided by the Nd:YAG laser is split by the beam splitter (BS) so part of the beam goes to the beam trap and is not used. The other part of the beam goes through the BS and impinges the first mirror (M 1 ). The fraction of the beam that traverses this mirror is attenuated and impinges the photodetector. Then, the photodetector generates an electric signal that is used as a trigger to the oscilloscope. The other part of the beam is directed by the M 1 mirror to another mirror, M 2, that directs the beam to a converging lens. The lens focuses the light but the beam homogenizer is placed after the focal point to avoid its damage, since the fluency can easily exceed 10 J/cm 2 on focus. The beam traverses the homogenizer and impinges the sample. On the other side of the sample there was placed a Mach-Zehnder interferometer to detect the ultrasonic waves in a direct transmission mode. In addition, the energy of the laser was decreased to its minimum value corresponding to 6 mj. After crossing the beam splitter the energy decreased to 3.4 mj and the fluency that impinged the sample was approximately 0.03 J/cm 2. The spot radius was approximately 2 mm and the pulse width was approximately 28 ns. In Fig. 4 it is shown the region tested with the lower fluency. Differently from [2], in this work it was not found a rectilinear grid of hot spots on the sample. No damage was found with this configuration and it was used to perform nondestructive inspection. The generation of longitudinal waves with a considerable amplitude using this configuration was achievable since the sample presents two layers, the first one is a transparent epoxy layer and the second one the carbon layer. This generates efficiently on the normal direction [3]. FIGURE 4. Region impinged by the laser with 0.03 J/cm 2, showing no damage. 937

4 FIGURE 5. Waveform of the ultrasonic wave. RESULTS Velocity Measurement The velocity of the longitudinal wave in the thickness direction was measured in a region without damage of the sample. The setup presented on Fig. 3 was used to generate and detect the ultrasound wave in the sample. The waveform acquired is shown in Fig. 5. Figure 5 shows, besides the waveform in time domain, the Short-time Fourier Transform (STFT), the Continuous Wavelet Transform (CWT) and the signal de-noised using the Stationary Wavelet Transform (SWT). The waveforms were used to measure the velocity of the longitudinal wave through the measurement of the time difference between the 3P and P peaks. The CWT was used to help finding the peak time position. The velocity measured for this material on the thickness direction was 3,215 mm/µs. Images The same setup was then used to make an inspection and to generate images of the composite sample. The sample was placed over xy translation stages in such a way that it was moved while the generation and detection beams were kept aligned. The sample was inspected in a region of 20 mm x 20 mm with 1 mm of step size. To each position, the Nd:YAG laser was driven with a repetition rate of 5 Hz and the signal was detected using the oscilloscope. Figure 6 shows the images generated through the longitudinal wave peak amplitude measurement. 938

5 FIGURE 6. Images generated using peak amplitude. FIGURE 7. Images generated using time-of-flight. Figure 7 shows the images generated through the longitudinal wave time-of-flight. In Figs. 6 and 7 one can see the impact damage, with an approximate size of 10mm x 10 mm, detected on the sample. The impact caused a delamination inside the sample and the laser ultrasonics system was able to detect it. The sample was also inspected using a C-scan equipment with a 15 MHz transducer. The entire sample was imaged and a step of 0.5 mm was used. In Fig. 8 the resulting images generated using the peak amplitude are shown. The images show the impact damage in the center of the image. 939

6 FIGURE 8. C-scan images generated using peak amplitude. In Fig. 9 the images generated using time-of-flight gates are shown. In these figures the defect was also detected in the center of the sample. FIGURE 9. C-scan images generated using time-of-flight. 940

7 The images generated using the laser ultrasonics system enabled the visualization of the internal damage. The conventional C-scan confirmed the damage found by the laser ultrasonics system. CONCLUSIONS The Nd:YAG laser did not cause ablation on the surface of the composite material due to the use of the beam splitter to reduce the laser energy and the use of the homogenizer to provide a better distribution of the laser energy over the spot cross section. This resulted on a lower fluency on the spot avoiding surface damage. C-scan imaging was accomplished to detect defects in the samples using the laser ultrasonics technique. These images were compared with those obtained using a commercial C-scan machine based in conventional ultrasonics transducers, showing that the laser ultrasonics system is capable to detect and image an internal defect without damaging the sample and without using a coupling gel or water bath on the composite material. The laser ultrasonics system can be improved to provide a better resolution image through the use of an automatic scanning xy table in order to acquire more data points. ACKNOWLEDGEMENTS The authors wish to thank Dr. Chiaki Miyasaka and the student Robert Cyphers. One of the authors (JMSS) acknowledges the Brazilian agencies CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico), for the provision of a scholarship under the process number /2007-8, and CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior), for the provision of an international scholarship under the process number 4697/08-1. REFERENCES 1. T. Stratoudaki, C. Edwards, S. Dixon and S. B. Palmer, Laser based ultrasound using different wavelengths for the inspection of composite materials, in Review of Progress in QNDE 21, edited by D. O. Thompson and D. E. Chimenti, AIP Conference Proceedings vol. 615, American Institute of Physics, Melville, NY (2002), pp C. Edwards, T. Stratoudaki, S. Dixon and S. B. Palmer, Laser based ultrasound generation efficiency in carbon fiber reinforced composites, in Review of Progress in QNDE 20, edited by D. O. Thompson and D. E. Chimenti, AIP Conference Proceedings vol. 557, American Institute of Physics, Melville, NY (2001), pp M. Dubois, F. Enguehard and L. Bertrand, A two-layer model for the laser generation of ultrasound in graphite-epoxy laminates, Review of Progress in QNDE 14, edited by D. O. Thompson and D. E. Chimenti, Plenum Press, New York, NY (1995), pp