EFFECT OF SURFACE COATINGS ON GENERATION OF LASER BASED ULTRASOUND

EFFECT OF SURFACE COATINGS ON GENERATION OF LASER BASED ULTRASOUND V.V. Shah, K. Balasubramaniam and J.P. Singh+ Department of Aerospace Engineering and Mechanics +Diagnostic Instrumentation and Analysis Laboratory Mississippi State University, MS 39762 INTRODUCTION Ultrasonic techniques are being currently used in a wide range of applications. However, there are situations where it is difficult to utilize ultrasound for NDT purposes. They are severely limited in non-contact applications due to attenuation in air. Since the present day technology requires a physical contact or a couplant between the transducer and the specimen surface to minimize diffraction effects, the temperatures sensitive piezoelectric transducers are not suited for hostile operating environments and extreme temperature gradients. Another drawback of piezoelectric transducers is the narrow banded source signals. Generation of very high frequency ultrasound becomes difficult and expensive. Complex contoured specimens are, also, difficult to handle, since traditional ultrasound is sensitive to the normalization of the incident ultrasonic beam. As with all mechanical devices, the rate of scanning of these transducers is very slow. Laser based ultrasound techniques provide an opportunity to make non-contact ultrasonic measurements and have a tremendous potential for several critical NDE applications. Laser based ultrasound offers very broad band frequency response. It is an important tool to study the true spectral interaction of longitudinal, shear and surface waves with defects since the signal is very wide-banded. The use of laser based non-contact interferometric detectors completely eliminate couplant problems. This also eliminates damping of surface motion as the physical contact between the detector and the specimen is absent. Optical scanning is fast compared to mechanical scanning and is more suitable for complex contours. Review of Progress in QuanJitative Nondestructive Evaluation. Vol. 14 Edited by D.O. Thompson and D.E. Chimenti. Plenum Press. New York. 1995 569

rn"r. E8--...!i 2 '" CEJ I '. ":." ' '" -." --....----------, r,,,, _ ff7 Figure 1. Experimental setup. Figure 2. Extraction of signal characteristics. PROBLEM STATEMENT In-service structures usually have protective corrosion resistant coatings. Surface coatings make the requirement of high laser power redundant for the generation of ultrasound. This is because the surface coatings decrease the reflectivity of the generation surface, which improves the energy absorption. Coatings also ablate at lower energy intensities than the structural materials. And hence they generate comparable amplitude of ultrasound as conventional piezoelectric generators. They also prevent ablation of the substrate material. The aim of this research is to understand whether these coatings would assist the generation of laser ultrasound. Earlier work has been reported studying this phenomena. This study furthers the earlier work using the frequency spectrum analysis and compares the results for three different frequencies of laser irradiation. The primary objective of this study was to experimentally observe the effect of coatings on the generation characteristics of ultrasonic waves. Another objective was to understand the generation characteristics of different frequencies of laser irradiation on the generated ultrasound characteristics. The directivity patterns of time and frequency domain features of laser generated ultrasonic longitudinal and shear waves were also studied. EXPERIMENTAL SETUP The experimental setup is as shown in the figure 1. An unfocussed Q-Switched Neodymium Y AG laser source was incident on the surface of the specimen. The 570

through-transmitted ultrasonic wave was received by piezoelectric ceramic crystal transducers. Table 1 describes the various surface coatings used, for different types of receivers, specimens and laser irradiation frequencies. The transducer signal was digitized using a Tektronix 320 digital oscilloscope and the digitized data was stored, for further processing, in a 486 PC through an IEEE488. 2 bus interface. The directivity tests were conducted on hemi-cylindrical steel specimen. The effect of different laser irradiation frequencies was carried out on a rectangular prismatic specimen. The time and frequency features were obtained through post processing of the digitized signal. The specimen surface did not show any damage due to the incident laser or coating residue at low energy levels. The directivity test were carried out on a hemi-cylindrical specimen 62 mm in radius. The laser beam was incident on the flat side of the specimen and the piezoelectric receivers were placed on the curved side of the specimen. The surface coatings were reapplied after every reading since there was ablation. Consistency in the thickness of the surface coatings were difficult to gauge and maintain. To minimize material and electrical noise, and variation in laser power density, an average of 16 signals was used per reading. The effect of different frequencies of laser irradiation on the generation of ultrasound was carried out on a rectangular prismatic steel specimen 18 mm in thickness. The three different frequencies of laser irradiation were produced using harmonic generators. Significant attenuation was observed in the blue light at 355 nm limiting the energy to 5 mj per pulse. Focussing of the laser beam was achieved using a mounted convex lens positioned at near focal length, 2/3 focal length and at 1/3 focal length. The nominal beam diameter of 9 mm was reduced to 6.1, 3.1 and 0.3 mm respectively. The coated surfaces were studied without focussing, as focussing significantly increased the ablation. The waveforms obtained from the experiments above were processed in a software to obtain a number of signal characteristics. Two types of features were extracted; the time domain envelope features and the frequency domain spectral features. Gating was used to isolate the ultrasound signals. The waveforms were normalized in their amplitude and low frequency noise was filtered out. The time domain envelope was obtained using a Hilbert transform. Figure 2 illustrates the features extracted. RESULTS AND DISCUSSION The figures 3(a-c) show the digitized waveforms received from the oscilloscope for a 10 MHz longitudinal receiver. These waveforms are obtained with the receiver placed parallel to the laser beam. Figures 4(a-c) show the polar directivity (0-90 ) plots of the absolute peak, for a 10 MHz longitudinal transducer. The x-axis is along the specimen surface, whereas the y-axis is normal to the surface. There is a significant increase in the amplitude of the longitudinal wave from an 571

2.0 4.0 48.0 (a) (b) 40.0 (c) :> :> :> 32.0 a 1.0 a 2.0 a 24.0.5.5.s 16.0 Q) Q) Q) -0-0 "0 8.0.. 1.s 0. 0.0 0.0 0. 0.0 0. -8.0.A., f' -16.0-24.0 9.9 10.4 10.9 11.4 11.9-2.0 9.9 10.4 10.9 11.4 11.9-32.0 9.9 10.4 10.9 11.4 11.9 Time in microseconds Time in microseconds Time in microseconds -1.0 L...o...L.-'----'.-l...LJ Figure 3. R-F signal obtained using a 10 MHz longitudinal receiver for a laser pulse energy of 8.2 mj. a) Uncoated control surface, b) Silicone coated surface, c) Mineral oil coated surface. uncoated surface to a coated surface. This can be attributed to the disintegration and ablation of the coating surface. With coating, the generated ultrasound becomes more directional(fig 4b & c). Figures 5(a-c) show the directivity test results for a 1 MHz shear receiver. The directivity tests show marked change in the amplitudes for the coated surfaces (fig 5b & c) as compared to the uncoated surface (fig 5a). The frequency spectrum of these waveforms also show distinct improvements(fig 6). Along with an increase in the peak frequency amplitude, the spectrum shows uniform increase in the bandwidth and 60.0. 50.0.s 40.0 ] 30.0 20.0 «10.0 0.0 -"'.L.L-.L.L-.L.L-.L.L-L.L.J 50.0.s 40.0 <I) ] 30.0 20.0 «10.0 50.0.s 40.0 <I) ] 30.0 20.0 «10.0 0.0-0.0 0.00.5 1.0 1.5 2.02.5 3.0 0.00.5 1.0 1.52.02.5 3.0 0.00.5 1.0 1.52.02.53.0 Amplitude in my. Amplitude in my. Amplitude in my. Figure 4. Polar directivity profiles showing peak amplitude obtained from the envelope of the R-F signal. The specimen surface corresponds to the x-axis. a) Uncoated control surface, b) Silicone coated surface, c) Mineral oil coated surface. 572

a.s.s.s G) G) G) "0 "0 "0 12.0 :> 12.0 12.0 a 6.0.g 6.0.g 6.0 'i5. a <r:: <r:: <r:: 0.0 0.0 0.0 0.0 6.0 12.0 0.0 6.0 12.0 0.0 6.0 12.0 Amplitude in m V. Amplitude in my. Amplitude in m V. Figure 5. Polar directivity profiles showing peak amplitude obtained from the envelope of the R-F signal for a 1 MHz shear receiver. The specimen surface corresponds to the x-axis. a) Uncoated control surface, b) Silicone coated surface, c) Mineral oil coated surface. significant increase in the energy under the frequency curve. A strong surface wave observed in the uncoated surface significantly decays in the coated surfaces. The observations of the effect of different frequencies of laser irradiation are discussed next. Tables 2 and 3 compare the time and frequency domain peak amplitudes in my, obtained for a 10 MHz longitudinal receiver. The comparison shows that although in general, the application of a coating surface increases the amplitude of the generated ultrasound, the amplification is dependent on the nature of the coating and the irradiation frequency. A higher amplitude was observed in the case of the blue and the green light as compared to that in case of infra red light. CONCLUSIONS It has been shown in this study that the signal to noise ratio of the generated laser based ultrasound signal, increases significantly in the longitudinal mode due to surface coatings. The directivity profile becomes directional towards the normal of the specimen surface. This may be due to the generation of ultrasound in the ablation regime. The frequency spectrum characteristics of the signals did not exhibit any detrimental effects due to the surface coating. The signals show a broader band-width for coated surfaces. The amplitude of the generated ultrasound depends strongly on the type of surface coating and the laser frequency. Coatings have shown to be detrimental to surface wave generation. 573

300.0 "0 ".s "S. "0 200.0 N " E 0 Z 100.0 - Uncoated control surface -- Silicone coated surface -- Canol a oil coated surface,l'-\ '\, i'l, I, i' I! Y I, I, /I I r" I,;,"I ' I I I " I! "V i\ \,,. ' \I II '\! 'I f. ' I '\1,! \' i ;1 I " ',,'I ", I, II ". ; il " 1,'1, If ; \,'.. f " '.I,,.' r 1.5 Frequency spectrum in MHz, ''''- Figure 6. Frequency spectrum for R-F signals obtained at 0 angle between the generation source and the receiver. Table 1: Experiment variables. Coating Types Laser Wavelength Receiver Modes Specimen Types Uncoated Surface Blue Light 1 MHz Shear (at 355 run) Silicone Lubricant Epoxy Paint Mineral Oil Green Light (at 532 run) 5 MHz Shear 5 MHz Longitudinal Canola Oil Infra-red Light 10 MHz (at 1064 run) Longitudinal Hemi-cylinder Steel specimen Rectangular prism steel specimen 574

Table 2: Peak amplitude in my, obtained from the envelope of the waveforms. These results were obtained for a 10 MHz longitudinal receiver. Energy Focussing Blue Light Green Light Infra Red SmJ/pulse (@ 3SS nm) (@ S32 nm) (@ 1064 nm) Unjocussed 0.07200 1.30000 1.44000 Uncoated Surface Focus 20162 1.60000 1.49000 2.72000 Focus 40162 2.00000 9.14000 5.04000 Focus 60162 4.32000 7.77000 6.64000 Silicone Unjocussed 48.0000 4.50000 7.20000 Mineral Oil Unjocussed 35.2000 56.0000 5.60000 Canola Oil Unjocussed 52.0000 31.3000 11.2000 Table 3: Peak amplitude obtained from FFT data, in my. These results were obtained for a 10 MHz L receiver. Energy Focussing Blue Light Green Light Infra Red SmJ/pulse (@ 3SS nm) (@ S32 nm) (@ 1064 nm) Unjocussed 0.067818 0.130855 0.135272 Uncoated Focus 20162 0.146773 0.157807 0.141380 Surface Focus 40162 0.214454 1.192157 0.158028 Focus 60162 0.432933 1.013020 0.209758 Silicone Unfocussed 7.292765 1.317116 0.208256 Mineral Oil Unfocussed 1.850960 13.80888 0.199831 Canola Oil Unjocussed 5.592371 8.068101 0.501375 575

REFERENCES 1. Birbaum, G. and White, G.S.," Laser techniques in NDE", Nondestructive Testing, Vol. 7, Academia Press, London, pp 259-365. 2. Fairand, B.P., et ai," Quantitative assessment of laser-induced stress waves generated at confined surfaces", Appl. Phys. Lett., Vol. 25, No.8, October 1974, pp 431-433. 3. Fox, J.A.," Effect of water and paint coatings on laser-irradiated targets", Appl. Phys. Lett., Vol. 24, No. 10, May 1974, pp 461-464. 4. Hutchins, D. A., Dewhurst, R. J. and Palmer, S. B.," Laser generated ultrasound at modified metal surfaces", Ultrasonics, Vol. 19, 1981, pp 103-108. 5. Ledbetter, H.M. and Moulder, J.C.," Laser-induced Rayleigh waves in aluminum", Letters to the ed., J.Acoust.Soc.Am., Vol. 65, No.3, March 1979, pp 840-842. 6. Monchalin, J.P.," Progress towards the application of laser-ultrasonics in industry", Review of Prog. in QNDE, Vol. 12, Ed. Thompson, D.O. and Chimenti, D.E., Plenum Press, New York, 1993, pp 495-506. 7. Scruby, C. B., Dewhurst, R. J., Hutchins, D. A. and Palmer, S. B.," Laser generation of ultrasound in metals", Research Techniques in Nondestructive Testing, Vol. 7, ed R. S. Sharpe, Academic Press, New York, 1982, pp 281-327. 8. von Gutfeld, R.J. and Melcher, R.L.," 20-MHz acoustic waves from pulsed thermoelastic expansions of constrained surfaces", Appl. Phys. Lett., Vol. 30, No.6, March 1977, pp 257-259. 576