Ultrasonic wave generation by time gated microwaves


Published on

Published in: Technology, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Ultrasonic wave generation by time gated microwaves

  1. 1. Ultrasonic wave generation by time-gated microwaves Bernard Hostena) ´ ´ Laboratoire de Mecanique Physique, Universite de Bordeaux I, URA C.N.R.S. No. 867, 351, ´ Cours de la Liberation, 33405-TALENCE Cedex, France Pierre Alain Bernard ´ ´ ` ´ Laboratoire de Modelisation Avancee des Systemes Thermiques et des Ecoulements Reels, ´ ´ Ecole Nationale Superieure de Chimie et de Physique de Bordeaux, Avenue Pey-Berland, B.P. 108, 33402 TALENCE Cedex, France Received 26 October 1997; accepted for publication 21 April 1998 This paper shows how the energy transfer between electromagnetic waves and elastic waves at the surface of fluid or solid materials could constitute a new way to generate ultrasonic waves without contact for the purpose of nondestructive evaluation and control. When time-gated microwaves strike the surface of a material, there is a generation of ultrasonic waves. The times-of-flight of these elastic waves inside the material decrease when the power of the incident microwaves increases. Therefore, the diminution of the propagation path indicates that the energy transformation appears in a volume close to the surface, the depth of which is increasing along with the power of the microwaves and depends on the material properties. Consequently, the elastic wave generation is attributed to the evolution of the power of the microwave during the gate width that governs the frequency content of the ultrasonic waves. The generation of ultrasound is checked in water and polymers. If the microwaves are not too attenuated in the material, the ultrasonic waves can be generated both at input air–solid and output solid–air interfaces. These peculiarities can generate new applications in the nondestructive evaluation and control of material. © 1998 Acoustical Society of America. S0001-4966 98 01308-3 PACS numbers: 43.35.Cg, 43.35.Mr, 43.35.Zc HEBINTRODUCTION as a tool to generate ultrasound without contact. The purpose of this paper is to show the conditions for which microwaves The nondestructive evaluation and testing NDE and can produce ultrasound at air–liquid, air–solid, and solid–airNDT of materials commonly involve the use of ultrasonic interfaces, and to describe initial experimental investigationswaves.1 The standard techniques utilize an immersion me- that will help to build the theoretical model of microwave-dium usually water or a coupling agent between the mate- elastic wave interaction.rial and the piezoelectric transducer. First we describe the characteristic of the time-gated mi- For over two decades, techniques that do not need any crowave source. Second, the generation in a volume close tocontact or coupling medium have been developed. The non- the air–liquid interface and the linearity of this phenomenoncontact techniques to generate ultrasound are based on the is discussed. It is shown that the ultrasonic wave frequencyelectromagnetic acoustic transducer EMAT air-coupled content is governed by the shape of the temporal gate. Third,transducers or laser impact.1 we introduce the generation of ultrasound by time-gated mi- Microwaves are frequently used in the nondestructive crowave near air–solid or solid–air interfaces, showing theinvestigation of materials, like x rays, to produce images of potential of this method for NDE and NDT. Finally, somethe transmitted intensity.1 Although the production of ultra- preliminary theoretical explanations are presented.sound from laser impact was intensely studied experimen-tally and theoretically, the microwave impact was not well I. MICROWAVE SOURCE DESCRIPTIONmentioned in the NDE community. Although R. M. White2 The electromagnetic wave is produced by a magnetronpresented the theoretical background in 1963, the production in the frequency range of 5.4–5.9 GHz C Band with a peakof elastic waves from modulated microwave energy was es- power of 1 MW. A pulse of 0.56 s is applied to the cathodesentially noticed for their physiologist effects. Human sub- of the magnetron. The repetition rate, 585 Hz, is fast enoughjects hear a ‘‘click’’ when the head is irradiated with high- to permit fast averaging if necessary. The wave is deliveredenergy microwave pulse. These effects were investigated by by a standard wave guide R48 of cross sectionthe microwave community.3,4 22.15 47.55 mm that imposes the propagation mode In 1984 R. L. Nasoni et al.5 presented some preliminary TE10 Fig. 1 . At these frequencies, the microwave wave-results on the generation of ultrasound from electromagnetic length is around 52 mm in vacuum and 62 mm in the wave-waves through immersed interfaces. To the authors’ knowl- guide. In order to use the lowest power possible and to proveedge, there have been no other attempts to use microwaves that compact sources could be designed in the future, a cou- pler was used to divide the power by 10. The amplitude ofa Electronic mail: hosten@lmp.u-bordeaux.fr the pulse sent to the magnetron was adjusted to furnish a860 J. Acoust. Soc. Am. 104 (2), Pt. 1, August 1998 0001-4966/98/104(2)/860/7/$15.00 © 1998 Acoustical Society of America 860
  2. 2. first transmission through the water layer is shown with two successive echoes. The times-of-flight are 44.19 and 44.22 s, respectively, between the first two echoes and the two subsequent ones. The time is measured with the cross- correlation function between echoes. The difference between the time measurement gives the order of the experimental errors that are mainly due to the slight shape difference of the echoes Fig. 2 . This observation proves that there is a transfer of energy between the time-gated microwave and ultrasonic waves near the air–liquid interface. B. Frequency domain The shape of the gate that modulates the microwave was acquired and is shown in Fig. 3 a and its spectrum in Fig. FIG. 1. Experimental arrangement to generate ultrasonic wave in water. 3 b . The bandwidth upper limit of this pulse is limited to around 1 MHz. Figure 4 compares the frequency domain ofpower at the end of the waveguide varying from 0 to 100 the first echo generated by the microwave and by the gen-kW. Another coupler diverted a small part of this energy to a erator in the pulse-echo mode. The frequency spectrum ofmilliwattmeter HP 436A in order to measure the micro- the ultrasonic wave generated by the microwave and filteredwave power delivered at the end of the waveguide. by the transducer receiver is comprised at 6 dB between 0.1 and 0.8 MHz. Then it seems reasonable to assume that the frequency content of the ultrasonic wave generated byII. EXPERIMENTAL VERIFICATION OF ULTRASONICWAVE GENERATION the microwave source is controlled by the shape of the gate. In the future, it will be more convenient to use a broadbandA. Generation at the air–liquid interface receiver, such as a laser probe. Figure 1 presents the apparatus used to observe the gen-eration of ultrasonic waves in water. Ultrasonic waves are C. Linearityreceived by an immersion transducer at a central frequencyof 1 MHz Panametrics V302 . The distance between the end Figure 5 presents the evolution of the peak amplitude ofof the microwave guide and the water surface is d2 20 mm . the first echo generated by the microwave function of theThe distance d1 33 mm is easily determined using the ul- instantaneous power of the microwave, in the range of 0 totrasonic transducer in pulse-echo mode and measuring the 100 kW. This corresponds to a mean power of 0 to 34 W.times-of-flight between echoes reflected by the water surface. The energy transformation between microwave and ultra-The time-of-flight between the two first echoes is 44.35 s sound is a linear phenomenon in this energy range. velocity in water 1.49 mm/ s at 22 °C . The amplification of the receiver was around 40 dB. Figure 2 presents the time history of the ultrasonic wave Clearly low power sources can be used in the future to gen-generated by the microwave at the interface air–water. The erate ultrasonic waves in liquids. FIG. 2. Ultrasonic waves generated by microwave at the air/water interface.861 J. Acoust. Soc. Am., Vol. 104, No. 2, Pt. 1, August 1998 B. Hosten and P. A. Bernard: Ultrasonic wave generation 861
  3. 3. FIG. 3. a Time history of the gate. b Frequency spectrum of the gate.D. Shift of the time-of-flight versus power III. GENERATION AT THE INTERFACE AIR–SOLID AND SOLID–AIR It is noticeable to consider time-of-flight as a function ofsource power. This time should be around 22.1 s according A. Weakly absorbing materialsto the previous measurement. This cannot be known abso- The experimental setup Fig. 1 is slightly modified tolutely because there is an uncertainty between the trigger investigate ultrasonic wave generation in a solid. A contactposition and the beginning of the magnetron oscillation. transducer Panametrics 101; Central frequency 0.5 MHz isHowever, the waveforms at various amounts of power can be connected to the solid through a coupling medium Fig. 7 .compared. As shown in Fig. 6, the times-of-flight decrease as The receiver is replaced by a more powerful amplifier Pana-the power increases. If a source region between microwave metrics 5058 PR; Gain: 60 dB since the ultrasonic waveand elastic wave is considered near the surface, one can as- amplitudes in the solid were much smaller than in water.sume that the source region depth is increasing with power Using the pulse-echo technique, the time-of-flight for aand the propagation distance of the ultrasonic wave is de- round trip in a block of PVC was found equal to 25.6 s.creased. This depth can be estimated to around 370 m 0.25 This value corresponds to a thickness of 29.5 mm with a s in water for 50 kW and is about 100 times smaller than longitudinal wave velocity of 2.3 mm/ s. Figure 8 presentsthe electromagnetic wavelength. the ultrasonic echoes generated by the microwave source at a It is important to notice that the oscilloscope was trig- power of 100 kW. Although the noise in this waveformgered by the gate, with no variation of the trigger position. could be easily suppressed by averaging, it is presented inHowever, if one expects a slight supplementary delay when this way to estimate the signal-to-noise ratio.the power is increased, the shift would be the reverse of what The time-of-flight of the first echo U 1 is about 12 s.it is shown in Fig. 6. This measurement is not accurate since the beginning of the echo is not well defined, nevertheless it proves that theFIG. 4. Modulus of the frequency spectrum. Solid line: first ultrasonic echogenerated by the microwave. Dashed line: first ultrasonic echo generated by FIG. 5. Peak amplitude of the first echo generated by the microwave versusthe generator in the pulse-echo mode. power.862 J. Acoust. Soc. Am., Vol. 104, No. 2, Pt. 1, August 1998 B. Hosten and P. A. Bernard: Ultrasonic wave generation 862
  4. 4. FIG. 6. First echoes generated by the electromagnetic wave for various values of the electromagnetic power.source is located near the upper surface of the plate. The sponds to an estimated source region depth of 1 mmmeasurement of the delay between echoes U 1 and U 2 is (velocity 1.7 mm/ s) which is much larger than in water.more accurate since it is possible to isolate each echo with a This must be explained by the attenuation of microwavestemporal window and to measure the delay by a standard which is much larger in water than in PVC.cross-correlation procedure. The result, 25.6 s, is identicalto the measure made with the pulse-echo technique. B. Reflecting materials Another echo L 1 appears between U 1 and U 2 . The ar-rival time of this echo, approximately 25 s, corresponds to As for optical waves, metals are almost pure reflectorsa round trip inside the plate. That means there is an impor- for electromagnetic waves. To observe the generation of ul-tant source of ultrasonic waves located near the lower sur- trasonic waves in metal, the PVC plate was replaced by aface. The delay 12.5 s between the echoes U 1 and L 1 is 49.3-mm-thick plate made of aluminum. The correspondingslightly less than the time-of-flight for one trip inside the waveform is not shown here because the ultrasonic echoesplate. Therefore the sources at upper and lower surfaces are were corrupted by in the ambient electromagnetic noise, butlocated in a volume with a small thickness. The accurate with a better setup one can imagine producing ultrasonicmeasurement of the source region depth must be performed waves in metals from electromagnetic waves. To enhance thewith a more precise setup since its value seems much smaller production of ultrasound, a 2-mm-thick water layer wasthan in the case of water. poured on the surface. The waveform is presented in Fig. 10. By using a more absorbing material, it is easier to reveal The ultrasound amplitude was so important that the gain ofthe presence of this source region. The PVC is replaced by a the receiver was set to 0 dB. The first arrival time1-cm-thick plate made of paraffin wax with 10% carbon ( 8.2 s) in the first echo is only slightly larger than onepowder. The waveforms, presented in Fig. 9, show the varia- trip in the aluminum plate ( 7.8 s). There is generation oftion of the amplitude and the time-of-flight of one echo as ultrasound everywhere in the water layer and almost at thea function of the electromagnetic wave power. As previ- water–aluminum interface.ously, the time-of-flight is decreased when the power isincreased. The difference is the order of 1 s which corre-FIG. 7. Experimental setup to generate and receive the ultrasonic wave in FIG. 8. Ultrasonic waves generated by microwave source at air/solid andsolid. solid/air interfaces electromagnetic power 100 kW .863 J. Acoust. Soc. Am., Vol. 104, No. 2, Pt. 1, August 1998 B. Hosten and P. A. Bernard: Ultrasonic wave generation 863
  5. 5. FIG. 9. Ultrasonic wave generated in a plate made of paraffin wax with 10% carbon powder. In order to explore the possible application to the in situ ii The transduction is linear and occurs in a zone nearnondestructive evaluation of metallic materials, a very thin the interface.layer of paint was sprayed on the aluminum surface. The iii The source region depth depends on the materialgain of the receiver was again increased to 60 dB. The wave- properties and the power of the electromagnetic wave.form in Fig. 11 shows the noise due to the electromagnetic The higher the power or the higher the absorption ofpulse and a first echo arriving at around 8 s. This result is the electromagnetic wave, the deeper the source re-promising for NDE applications since structures are often gion. Its depth is much smaller than the penetrationpainted. depth of the microwave inside the material, since the microwave wavelength is much larger than the source region.IV. MODEL Laser impacts also create elastic waves via several To build a model for the interaction between electro- mechanisms: radiation pressure, electrostriction, Brillouinmagnetic waves and elastic waves more precise experiments scattering, and thermoelastic expansion. The latter is domi-and investigations are necessary. However, the following ex- nant under normal circumstances. These mechanisms andperimental considerations will be useful to develop a model: their applications are completely described in literature andi The evolution of the electromagnetic wave power cre- textbooks.6,7 In addition, the presence of a ‘‘precursor’’ was ates elastic waves in a frequency domain imposed by observed in laser generation and explained with one- or two- the gate width. dimensional models,2,8,9 taking into account the presence of a FIG. 10. Ultrasonic wave generated at the water/aluminum surface electromagnetic power 100 kW .864 J. Acoust. Soc. Am., Vol. 104, No. 2, Pt. 1, August 1998 B. Hosten and P. A. Bernard: Ultrasonic wave generation 864
  6. 6. FIG. 11. Ultrasonic wave generated at the paint/aluminum surface electromagnetic power 100 kW .source buried below the surface. The ‘‘precursor’’ appear- V. CONCLUSIONSance is equivalent to what is observed from the evolution of This paper highlights the transformation of energy be-times-of-flight Fig. 6 . Since the electromagnetic source is tween electromagnetic waves and elastic waves through thelarge in comparison with the ultrasonic wavelength, it is con- surface of liquid or solid materials.ceivable that the one-dimensional model would be The experimental considerations lead to a one-appropriate.2,8,9 dimensional thermoelastic model. More experiments must be As explained in Ref. 9, a nonfocused laser source in the conducted to link the electromagnetic and elastic materialthermoelastic regime generates two waves located at both properties to the elastic waves characteristics. These experi-faces of the plate. A similar analysis would be appropriate mental observations will permit us to model the energy trans-for a microwave source. The stress induced by a tempera- formation in order to estimate the penetration depth, to pre-ture rise T is given by dict the response of any material and to consider the u x,t applications to the nondestructive evaluation of materials. x,t C T, 1 Ultrasonic waves can be generated at interface air–solid x and also solid–air. If the electromagnetic attenuation is notwhere u is the normal displacement in the ultrasonic field, C too significant, the electromagnetic wave can generate ultra-is the diagonal component of the rigidity tensor in the x sound inside the materials in a zone located near a cavity ordirection normal to the interface, and is the thermal stress a delamination. Clearly, some important applications to thecoefficient. The interfaces at x 0 and x L are stress free, nondestructive testing of material can be accomplished.then (0,t) 0 and (L,t) 0. Equation 1 implies there is There is a very large spectrum of applications since thea displacement gradient at x 0 and x L. Therefore, two gate width and shape of the microwave can be controlled towaves are produced at both interfaces if the material does not produce lower frequency content. In the other way, ultra-absorb too much of the electromagnetic energy. If the ab- sonic waves with very high frequency content can be pro-sorption is negligible, the spatial dependence of T is weak, duced with shorter pulses and even a monocycle pulseand the second term on the right-hand side in the wave equa- source.10tion 2 2 ACKNOWLEDGMENTS u x,t u x,t T C 2 The authors gratefully acknowledge the CEL Centre t2 x2 x d’Essais des Landes for supplying the time-gated micro-can be neglected and the waves are produced near the inter- wave source and the help of Bernard Desvergnes, Jean Bran-faces. If the absorption increases, this term is responsible for dier, Jean-Claude Devant and Stephane Lallement.the buried sources of ultrasounds and the depth of the gen-eration zone increases. 1 Proceedings of Review of Progress in Quantitative NonDestructive Evalu- This model seems even more appropriate for the micro- ation, edited by D. O. Thompson and D. E. Chimenti Plenum, Newwaves than for the optical waves since their wavelengths are York , Vols. 1–16. 2much larger. For instance, in this paper the microwave wave- R. M. White, ‘‘Generation of Elastic Waves by Transient Surface Heat- ing,’’ J. Appl. Phys. 34„12…, 3559–3567 1963 .length is larger than the thickness of the tested materials. 3 J. C. Lin, ‘‘Further Studies on the Microwaves Auditory Effect,’’ IEEEComparison between theory and experiment will be pre- Trans. Microwave Theory Tech. MTT-25„11…, 939–943 1977 . 4sented in a later paper. D. Borth, ‘‘Theoretical Analysis of Acoustic Signal Generation in Mate-865 J. Acoust. Soc. Am., Vol. 104, No. 2, Pt. 1, August 1998 B. Hosten and P. A. Bernard: Ultrasonic wave generation 865
  7. 7. rials Irradiated with Microwave Energy,’’ IEEE Trans. Microwave Theory on laser-generated ultrasound,’’ J. Acoust. Soc. Am. 88, 1494–1502 Tech. MTT-25„11…, 945–953 1977 . 1990 .5 R. L. Nasoni, G. A. Evanoff, P. G. Halverson, and T. Bowen, ‘‘Thermoa- 9 ´ A. Henault, A. Cournoyer, F. Enguehard, and J. Bertrand, ‘‘A study of coustic Emission by Deeply Penetrating Microwave Radiation,’’ in IEEE dynamic thermal expansion using a laser-generated ultrasound 1-d Ultras. Symp. 1984 , pp. 633–638.6 model,’’ Proceedings of the 9th International Conference on Photoacous- D. A. Hutchins, ‘‘Ultrasonic generation by pulsed laser,’’ Physical Acous- tics, edited by W. P. Mason and R. N. Thurston New York, 1988 , Vol. tic and Photothermal Phenomena, edited by S. Y. Zhang, Nanjing, China XVIII. 1996 , pp. 370–374. 107 C. B. Scruby and L. E. Drain, Laser Ultrasonics Techniques and Appli- F. C. Chen and W. C. Chew, ‘‘An impulse radar nondestructive evaluation cations Hilger, Bristol, 1990 . system,’’ in Review of Progress in QNDE Plenum, New York, 1997 ,8 K. L. Telschow and R. J. Conant, ‘‘Optical and thermal parameter effects Vol. 16A, pp. 709–715.866 J. Acoust. Soc. Am., Vol. 104, No. 2, Pt. 1, August 1998 B. Hosten and P. A. Bernard: Ultrasonic wave generation 866