A Photophone-Based Remote Nondestructive Testing Approach To

Home , Photophone, Telephony

Original Article Structural Health Monitoring 2018, Vol. 17(2) 135–144 Ó The Author(s) 2017 A photophone-based remote Reprints and permissions: sagepub.co.uk/journalsPermissions.nav nondestructive testing approach to DOI: 10.1177/1475921716686772 interfacial defect detection in journals.sagepub.com/home/shm fiber-reinforced polymer-bonded systems

Tin Kei Cheng1 and Denvid Lau1,2

Abstract Externally bonded fiber-reinforced polymer is an increasingly popular material to be used in strengthening and retrofitting aging structures. In such structures, debonding defects may occur at or near the interface between fiber-reinforced polymer and concrete. As such debonding in fiber-reinforced polymer-bonded systems is generally brittle in nature, there is a need of a reliable inspection technique that can provide early warning of interfacial defects such that premature failure of fiber-reinforced polymer-strengthened structures can be avoided. A remote nondestructive testing approach based on the working principle of a photophone is presented here as an economical alternative to laser Doppler vibrometry for detecting interfacial defects. Concrete specimens retrofitted with fiber-reinforced polymer are excited acoustically by white noise, while the surface of the structure is illuminated by a light source. If an interfacial defect exists beneath the surface, the surface will exhibit a frequency response different from an intact surface. The surface of the fiber-reinforced polymer portrays the role of flexible mirror in a photophone, which encodes information about surface vibration into amplitude- modulated light signal. A light detector then captures the irradiance of the reflected beam, and the amplitude modulation is converted into frequency domain in post-processing. With this technique, defect dimensions and thus damage extent can be inferred from the frequency spectrum obtained. The obtained results correspond well with the theoretical calcula- tion, demonstrating the robustness and the applicability of the proposed technique in civil infrastructure.

Keywords Acoustic-laser, amplitude-modulated modal analysis, fiber-reinforced polymer, interfacial defect, nondestructive testing, photophone

Introduction good environmental resistance, and excellent resistance against fatigue. Because externally bonded FRP becomes The maintenance of aging structures is a challenging issue 1 more popular in retrofitted structures, the failure modes faced by many nations. In most cases, strengthening and of FRP-bonded systems have thus been studied exten- retrofitting of these deteriorated structures may be more sively to better understand the mechanical behavior of economical than reconstruction, especially in situations such composite structures.2–6 Failures in FRP-bonded involving the preservation of historic sites. The use of reinforced concrete (RC) beams can be classified into fiber-reinforced polymer (FRP) composites as an externally bonded element to retrofit structural elements such 1 as beams, columns, slabs, and bridge decks in order to Department of Architecture and Civil Engineering, City University of Hong Kong, Hong Kong restore the design capacity and structural redundancy of 2Department of Civil and Environmental Engineering, Massachusetts deteriorated structures has emerged as a popular Institute of Technology, Cambridge, MA, USA strengthening and retrofitting technique. FRP, as a thin sheet of composite material, has a lot of desirable proper- Corresponding author: ties when compared with the conventional bulk construc- Denvid Lau, Department of Architecture and Civil Engineering, City University of Hong Kong, Hong Kong. tion materials, such as high strength-to-weight ratio, Email: [email protected] 136 Structural Health Monitoring 17(2) two major mechanisms: (1) flexural failures of critical sec- tions, such as crushing of concrete and FRP rupture and (2) debonding of FRP from the RC beams.2–6 The near- surface debonding defects resulting from the latter case will be referred to as interfacial defects in this article, and they are the focus of this study. Interfacial defects often occur in regions with high stress concentration and/or shear stress discontinuity, which are positivelycorrelated with (1) regions of material discontinuity, (2) poor instal- lation including inadequate surface preparation of the substrate, and (3) the presence of cracks (e.g. propagation of flexural or flexural-shear cracks from the concrete sub- Figure 1. A representative arrangement of the photophone.26 strate to the FRP-concrete interface, resulting in delami- A beam of sunlight is concentrated upon a flexible mirror on the nation). The failures modes due to interfacial defects are transmitter side. After reflection, the beam is again made brittle in nature and will result in premature failures of parallel by means of another lens. The beam is then received at FRP-strengthened elements if not promptly resolved.2–6 a distant location upon a parabolic reflector, the focus of which is placed on a selenium photoresistor, connected in a local Current nondestructive testing (NDT) approaches circuit with a battery and telephone. developed for civil engineering application include the use of ultrasound, X-ray and neutron radiography, near-field and far-field radar, and infrared thermogra- required to assemble the NDT device put forward in phy.7–19 However, these techniques may not always be this article can readily be purchased from consumer optimal for the remote detection of interfacial defects.1 electronics retailers, and the total cost of a functioning Passive thermography typically has insufficient spatial setup is around a hundredth of the original work. 18 Originally conceived by Alexander Graham Bell who resolution to resolve small interfacial defects. 25 Radiography involves the use of hazardous ionizing invented the telephone in 1880, a typical photophone 10,11 concentrates sunlight onto a flexible mirror which radiation. The radar approach may have issues if 26,27 carbon fiber–reinforced polymer (CFRP) sheets being reflects the sunlight to a remote receiver (Figure 1). penetrated by radar are not arranged in a parallel Excited by the speaker’s voice, the flexible mirror mod- orientation as CFRP is a highly conductive material at ulates the reflected sunlight’s radiant intensity by flex- microwave frequencies.14 Finally, while near-field radar ing convex and concave according to the changing and ultrasound are promising approaches, they require sound pressure, dispersing, and converging the beam of a close proximity to the subject.14,17,19 Recently, high- sunlight. This amplitude-modulated radiant intensity is speed video in conjunction with motion magnification converted back to an audio signal when detected by the has been applied to remotely evaluate modal behavior. receiver. The photophone eventually achieved the first Despite being promising in detecting larger scale dam- wireless telephone communication, predating the first ages, as of the moment of writing, the maximum fre- spoken radio transmission by almost two decades and quency measurement demonstrated by the high-speed laying the groundwork for fiber-optic communication a video technique is 2500 Hz, which is below the typical century later. Although considered by Bell to be one of fundamental frequencies of interfacial defects as his greatest achievements, even more superior than the explored in this article.20 To complement existing NDT telephone, the photophone never went into mainstream 25,28,29 approaches, an acoustic-laser technique has been usage and it is found mostly in military use. devised specifically to detect interfacial defects from a Nowadays, the photophone is commonly known as the distance based on their vibrational behavior.21–24 It has ‘‘spy microphone’’ in hobbyist projects, with few real- been shown that interfacial defects in FRP-bonded sys- life applications. In this article, the interest in photo- tems can be modeled as a combination of a clamped phone is renewed by considering an NDT application plate (representing the FRP surface) and an acoustic in the detection of interfacial defects in FRP-bonded cavity (representing the air gap at the interface), resem- systems. The working principle of the modified photo- bling the physics of a drum. In such defects, the funda- phone will first be introduced with the relation to previ- mental mode of vibration lies typically at the order of ous acoustic-laser studies, followed by experiments with kHz.21 However, the dedicated setup used in previous artificially induced delamination-like interfacial defects acoustic-laser studies may cost a million US dollars on FRP-bonded concrete specimens. Finally, benefits and might prove an obstacle to the commercialization and limitations of this photophone-based technique will of the acoustic-laser technique. This study presents an be addressed based on experimental results, and future economical augmentation to the previous works based research direction including an extension of this tech- on the design of a photophone. Electronic parts nique to general modal analysis will be discussed. Cheng and Lau 137

Figure 2. An illustration of the photophone-based experiment setup. (a) An FRP-bonded concrete slab with artificially induced interfacial defects is excited acoustically by white noise, while a point on the surface of the structure is illuminated by a laser. As specular reflection is strong in our specimens, a lens system is not installed in the light detector, though a collimator is used to isolate low-frequency noise due to ambient lighting. (b) The surface of the FRP portrays the role of flexible mirror in a photophone, alternately focusing and diverging the reflected light, encoding information about surface mechanical vibration into amplitude- modulated light signal, which is captured by a light detector. (c) The amplitude modulation of irradiance is converted into frequency domain in post-processing. If an interfacial defect exists beneath the surface, the surface will exhibit a frequency response different from an intact surface. Defect dimensions and damage extent can be inferred from the obtained frequency spectrum.

Working principle However, the technique measures only surface vibra- tions. Therefore, to obtain a holistic view of the struc- The acoustic-laser technique ture’s health, the acoustic-laser technique must be Although previous acoustic-laser studies focus on combined with other NDT techniques to reveal subsur- detecting debonding defects located at or near the interface defects. face between a sheet of FRP and its substrate, it is in fact a general NDT method for remote evaluation of structures that can be excited acoustically and would The photophone-based approach and experiment induce a vibration on an exposed measurable sur- setup 21,30 face. Defects are identified based on the frequency In this study, the measurement setup is refined to response of the surface vibration. Previous works demonstrate the flexibility in deployment and commer- involve exciting the FRP-bonded structure by a fre- cial viability of the acoustic-laser technique quency sweep from a high-power standoff parametric (Figure 2(a)). An ordinary loudspeaker is used in place acoustic array (PAA), and the resulting vibration is of the PAA as the acoustic excitation source given that monitored remotely by a laser Doppler vibrometer the fundamental modes of interfacial defects have been (LDV) with the peaks in the frequency response spec- 21 30 shown to be in the order of kHz, thus can be excited trum corresponding to vibration modes of the defect. by consumer-grade electronics. White noise is intro- The benefits of this design include the ability of long- duced for acoustic excitation instead of a frequency range measurements up to 30 m, the robustness against sweep to cover a wide range of frequencies in a short environmental noise as the technique is based on inter- period of time. The LDV is replaced by the low-cost ference, and the ability to connect measurement result photophone-based measurement system proposed in (surface vibration velocity) with defect characteristics this article, exchanging frequency modulation (FM) in for intuitive interpretation of experiment results. an LDV system with amplitude modulation (AM), with 138 Structural Health Monitoring 17(2) a tradeoff in accuracy and precision. Although the speed of sound in air; l, m, and n as positive integers of setup in previous works involving PAA and LDV pro- vibration mode numbers along the x, y, and z axes of a duces highly precise results, the cost of the setup is over Cartesian coordinate system with the x and y axes a million dollars. At around a hundredth of the original oriented along the orthogonal edges of a rectangular cost, the current setup provides comparable perfor- defect. For a square isotropic defect, its area can be mance, albeit at the expense of the resistance to envi- estimated by ronmental noise, as expected in an AM system. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Although the photophone was intended for voice ll;mh E 1:65h E transmission, it is realized that the transmission carries A = 2 ’ 2 ð1Þ vl;m 12rð1 v Þ f1;1 rð1 v Þ not only the voice but also information regarding the vibrational characteristics of the flexible mirror. Through where ll;m is a numerically calculated constant with white noise excitation or a frequency sweep, the modes l1;1’35:99. A table of ll;m for higher modes can be of resonance of the mirror can be probed. Furthermore, found in the work by Leissa.31 However, since FRP is it is noted that all reflective surfaces can act as the flex- intrinsically a transversely isotropic material, the quali- ible mirror, allowing modal analysis on a general object tative behavior rather than the exact outcome of the based on the principle of a photophone. If a surface is above relation should be noted. The depth of the defect dull, a retroreflective sticker or reflective paint can be can be identified by the following relation applied to the surface to enhance reflection. If a surface is rough, a lens system can be included to focus the dif- nvair 170 Lz = ’ ð2Þ fuse light reflected from the vibrating surface. 2fn f1 The working principle of our photophone-based method is illustrated in Figure 2. The specimen is It should be noted that the latter relation is linear in excited acoustically by white noise and is simultane- its harmonics, and thus can be decoupled from the frequency response of equation (1). However, as noted by ously illuminated by a beam of laser for vibration mea- 21 surement at a single point (Figure 2(a)). If a structure is Cheng and Lau, the vibration magnitude of the acoustically ideal in an open area and it is excited by acoustic cavity is generally one to two orders of magni- white noise, the resulting vibration would have equal tude weaker than the vibration of plate, as the sur- energy at every frequency bin, reproducing the incident roundings of the cavity (FRP, epoxy, and concrete) are white noise. However, in reality, all objects have a set much stiffer and denser than air. of resonance modes, and therefore, the frequency Although a laser is used for single-point measure- response is not flat. To be precise, spikes in the fre- ment, the efficiency would be low when an area is to be quency response correspond to resonance frequencies probed and the location of the defect is not known in of the object, save for a minor shift due to the effect of advance. In this scenario, the FRP surface could be damping. The vibration of the surface results in the illuminated by sunlight, a light emitting diode (LED) array, or equivalent even lighting setup (Figure 3(a)). periodic variation of the surface normal, which in turn Focusing the reflected rays with a lens system, the sin- causes the reflected beam of light to converge and gle photodiode detector can be replaced by a light sen- diverge at the same frequency as the vibration, resulting sor array, such as a high-speed camera, to obtain an in an amplitude-modulated beam of reflection carrying image of vibration activity over the entire area under information about the surface vibration (Figure 2(b) inspection. Significant response peaks observed in the and (c)). The reflected light is picked up by a photo- spectrum would indicate existence of defects (Figure diode circuit, amplified, and collected by a data logger 3(b)), prompting further investigation using the single- as voltage signal. The frequency-domain view of the point laser technique to localize the defect. It should be modulated signal is obtained via fast Fourier transform noted that in principle the illumination source is not (FFT), and vibration modes are identified by locating limited to visible light but applicable to the entire elec- response peaks in a plot of the frequency spectrum. tromagnetic (EM) wave spectrum. Defect dimensions and damage extent are inferred from the frequency spectrum as interfacial defects can be modeled as a clamped plate with an acoustic cavity Experiment underneath.21,31 In the following equation, the variable E is designated as Young’s modulus, n as Poisson’s Specimens ratio, r as volumetric density, h as thickness of the Three square concrete slabs with edge lengths of FRP-epoxy composite layer, and D = Eh3=12ð1 v2Þ as roughly 300 mm are casted with various interfacial the bending stiffness. Furthermore, A is defined as the defects manifested as near-surface voids induced by defect area; Lz as the depth of the air gap; vair as the inserting Styrofoams onto the mold of the concrete Cheng and Lau 139

Figure 4. FRP-bonded concrete slab specimens with artificial interfacial defects are shown here. The FRP is bonded to the surface facing the reader but is omitted in this illustration to reveal the underlying defects. Interfacial defects of three types of area: 25 3 25, 37.5 3 37.5, and 50 3 50 mm2 are investigated, together with five types of depths: 5, 10, 20, 30, and 40 mm.

Procedure A white noise recording with a sample rate of 48 kHz is played over a period of 11 s at 102.1 6 0.2 dB (sound pressure level (SPL)), while the environmental noise is fluctuating around 65–71 dB (SPL). Two control recordings without acoustic excitation of the same duration are recorded before and after the experiment to identify the influence of environmental noise. Over the entire period of recording, the FRP surface under inspection is illuminated by either a beam of 532 nm laser at a power of 80 mW or direct sunlight. When a laser is used for illumination, the source of acoustic excitation, a studio monitor (Yamaha MSP7 Studio), is placed 0.96 6 0.01 m away from the specimen surface at an angle + 15° 6 5° from the surface normal, while Figure 3. Simultaneous detection of multiple defects over a the laser and light detector are placed 1.80 6 0.01 and large area. (a) A broad-beamed light source replaces the laser in 1.7 6 0.1 m, respectively, away from the specimen and Figure 2 as the illumination source. Focusing the reflected rays located at + 25° 6 1° and 220° 6 10° on either side of with a lens system, the single photodiode detector can be the specimen. The position of the light detector has a replaced by a light sensor array to obtain an image of vibration larger variation as the reflection path is dependent on activity over the entire area under inspection. (b) Vibration the specimen surface normal. A laser line filter and col- signatures of multiple defects can be captured. limator are required as the light detector can easily be saturated by the 100 Hz or 120-Hz frequency of ambi- (Figure 4). The use of air voids is a common strategy to ent lighting (50 or 60 Hz mains electrical supply recti- emulate interfacial defects.7,8,14–16,30 The specimens are fied), which would have limited indoor use. Under cured for approximately 28 days in a water tank and sunlight, the studio monitor and the light detector are then dried at 50°C in a oven for 3 days. The surfaces of placed 1.0 6 0.1 m in front of the three specimens. A the specimens are sanded to reduce surface roughness photodiode connected to an amplification circuit is and are bonded to three layers of CFRP (TORAY placed in the path of the reflected light ray from the spe- UT70-30) using epoxy (SikadurÒ-300). The bonded cimens. The data are recorded on a computer via a data specimens are further allowed to cure for a minimum logger system operating at a sample rate of 50 kHz, giv- of 7 days before measurements are made. ing a Nyquist frequency of 25 kHz. In post-processing, 140 Structural Health Monitoring 17(2) the data are converted into frequency domain using together even though the scale of signal magnitude var- FFT of size 1024, multiplied by Hanning window, and ies across specimens. As this photophone-based tech- averaged using Welch’s method with an overlapping nique depends on measured irradiance, and the voltage region of 50%. The FFT size is chosen to optimize output of the photodiode depends heavily on the reflec- smoothness of the resulting spectrum rather than maxi- tivity of the measured surface and the positioning of mizing signal-to-noise ratio (SNR), as the latter would the light detector. Thus, reproducing the result in terms result in a highly jagged curve, a well-known artifact of of absolute irradiance is difficult to achieve as the mea- FFT. As the specimen is small in scale, the concrete slab sured specimen is not a smooth surface (i.e. an FRP has a high fundamental mode of vibration around fabric consisting of interweaving fibers). Although 2 kHz, resulting in low-frequency noise during the mea- repeatability in terms of measured voltage is degraded surement. Furthermore, aliasing toward the high- and the noise floor is not constant across measure- frequency end of the frequency spectrum is strong. ments, the clarity and consistency in the frequency Therefore, on each end of the frequency spectrum, a domain are high. In Figure 5(d), the frequency of each 2.5 kHz portion was cropped from the result. peak in the cluster is plotted as a ratio to the peak with the lowest frequency, further visualizing the effect of the clustering of peaks and implying that the distribu- Results tion of peaks is not random. Taking the peak with the The measurement results from single-point laser mea- lowest frequency to be the fundamental mode of vibration, it is plotted against the inverse of defect area surement with white noise excitation are shown in 2 Figure 5. The measurements of 50 3 50, 37.5 3 37.5, (Figure 5(e)). A linear relationship with R = 0.971 is and 25 3 25 mm2 defects are plotted in Figure 5(a) to observed, indicating strong correspondence to plate (c), respectively. Each solid curve represents a measure- theory. However, as the FRP, epoxy, and concrete are ment of a different defect specimen, while the dotted all much stiffer and denser than the trapped pocket of line represents the average of all frequency responses of air at the interface, the SNR is too low for defect depth each defect area. The response peaks appear to cluster to be decoupled from the current measurement.

Figure 5. Measurement results using the photophone-based technique. (a–c) The frequency response of 50 3 50, 37.5 3 37.5, and 25 3 25 mm2 defects, respectively, are shown. The behavior of each specimen is shown in the background as translucent curves, and the averaged response for each set of defect areas is displayed as a dotted line. (d) The frequency of each peak in the cluster is plotted as a ratio to the fundamental mode, further visualizing the effect of this clustering and implying that the distribution of peaks is not random. (e) The peak with the lowest frequency is taken to be the fundamental mode of the defect and is plotted against the inverse of defect area for obtaining a linear relationship with R2 = 0.971. Cheng and Lau 141

Figure 6. (a) Variation of frequency response under different magnitudes of acoustic excitation. Generally, a stronger excitation will result in a larger response but the increase in SNR is not linear. (b) The change in the direction of reflection (gray line) of an incident laser (green line) on a rough surface with small displacement.

Figure 7. (a) Change of frequency as a result of varying In Figure 6, the magnitude of acoustic excitation is excitation duration. (b) A magnified section to better visualize varied. Although generally a stronger excitation will the variation. It is observed that the longer the excitation, the result in a larger response, the increase in signal smoother the observed response. However, even though the strength is not linear. Occasionally, a higher mode may duration is increased exponentially, the signal strength is only have a stronger response than the fundamental mode increased marginally. (e.g. the third mode in Figure 6(a)). This observation is due to the fact that even though the photophone-based latter is only a loose upper bound as the excitation and technique has a high correlation with the specimen’s recording are manually synchronized in this study; resonance modes, it is an indirect measurement of the results below 0.2 s can be significantly influenced by specimen’s vibration. A mode that results in the highest human error. contrast in its reflection of the illumination source to In Figure 8, the measurement of the specimens illu- the light detector is favored. When a flat surface is dis- minated by sunlight is shown. The excitation duration placed (e.g. due to vibration), the direction of reflection is reduced to 5 s, and the response from a frequency remains unchanged. However, as the FRP surface is sweep from 100 to 20,000 Hz is also shown for compar- rough, the displacement of the surface amplifies the ison. Resonance modes of various defects are found to variation of the reflected light intensity, which may have superimposed on each other. However, a minor explain the nonlinear increase in SNR (Figure 6(b)). frequency offset of around 500 Hz is observed, which is It is also possible that the laser fall in between two car- possibly due to the specimens being clamped differently bon fibers, and the direction of reflection is erratic from the indoors single-point laser measurement, (Figure 6(b)). thereby altering the damping coefficient of the system. In Figure 7, the excitation duration is varied. It is Moreover, it is found that white noise excitation results observed that the longer the excitation, the smoother in a higher SNR than using frequency sweep, justifying the response. In the particular signal processing method the use of white noise for excitation in place of fre- used in this study (i.e. Welch’s method), however, even quency sweep. though the duration is increased exponentially, the signal strength only improved marginally. Judging from the result, an exposure of 2.6 s is optimal for manual Discussion trigger, while an excitation duration below 0.2 s is pos- From the consistency in inferring defect dimensions, sible for computer trigger. It should be noted that the the capacity of the measurement technique introduced 142 Structural Health Monitoring 17(2)

been made to combine different NDT techniques to offer a more comprehensive understanding of structural health,8,9 however, the techniques utilized remain penetrating in nature. As the design of the photophone-based technique is in an early stage, multiple improvements are possible for future studies. Due to the simplicity of the design, the setup can be further miniaturized as an all-in-one handheld device. If sunlight is used for illumination, the device can be plugged into a smartphone via its microphone input as the signal is compatible by design, with the defect being excited via the smartphone’s speaker. The frequency spectrum can be obtained via Figure 8. Measurement of multiple defects under plain sunlight existing audio-processing software with modal identifi- illumination. Resonance modes of various defects are found to cation in real time. A prototype has already been have superimposed on each other. Moreover, it is found that demonstrated while still using laser for illumination white noise excitation results in a higher SNR than using frequency sweep, justifying the use of white noise for excitation and is capable of providing a comparable level of mea- in place of frequency sweep. surement (Figure 9). If a stabilized drone is available, the device can be mounted on the drone for automated airborne areal measurement. Environmental noise such in this article as a low-cost alternative to the original as highway traffic can also be considered as a source of acoustic-laser technique is validated. Although only acoustic excitation. Moreover, noting that FRP is a square interfacial defects are investigated in this study, thin material, the loudspeaker or PAA can in fact be it is anticipated that a similar relationship between reso- omitted in the setup and replaced by a modulating illu- nance frequencies and defect dimensions would exist mination source, utilizing photoacoustic effect to gener- for other defect shapes. Indeed, supported by the evi- ate mechanical vibration from a long distance with 26 dence that known analytical solutions to other defect minimal dispersion. In this scenario, from the experi- shapes (e.g. circular, triangular) behave similarly,31 it ence of war-time development with the aid of high- has been suggested empirically that an inverse relation- powered light sources, the device is potentially capable 28,29 ship exists for all regular polygon-shaped interfacial of NDT over a range of kilometers. As a further defects between resonance frequencies and defect area, improvement, instead of a single light detector, the which differs only by a shape-dependent factor.21 FRP surface of interest can be focused by a lens system From the review of existing remote NDT approaches onto a detector array to obtain a two-dimensional (2D) in the introduction of this article, it can be seen that vibration map of the entire surface, visualizing the many of such techniques penetrate the specimen. In mode shape without the need of a scanning laser system radiography and infrared thermography, the resulting as in a scanning LDV. Similarly, the recording of a image contains information regarding the entire depth high-speed camera can be processed by treating each of the structure; in radar imaging, although the penetra- pixel as an independent light detector for the same pur- tion depth can be reduced by adjusting the frequency of pose, or the entire video is processed collectively using 32–34 the radar, the measurement still covers a certain depth the technique of motion magnification. of the structure. As concrete is a material containing It is envisioned that this photophone-based tech- coarse aggregates, the resulting scattering would most nique can be applied to a wider range of NDT applica- certainly result in a degradation in such techniques’ res- tions, including general modal analysis. The current olution. It may thus be difficult to distinguish a subsur- application of the technique in FRP, which is a fabric face defect from an interfacial defect, which has material, is already one of the worst possible scenarios different failure modes, if access is restricted to the ante- for technologies based on surface reflection. Although rior surface (e.g. when measuring a wall). Fortunately, the absolute comparison across results measured in dif- the acoustic-laser technique, the current modification ferent conditions is challenging for FRP surfaces, the inclusive, is based entirely on surface reflection. This robustness and consistency in frequency domain are ensures that the measured response is predominantly high. Hence, this technique is most suited for measure- due to interfacial defects. Therefore, a full picture of ments focusing on the frequency domain. In retrospect, both interfacial and subsurface defects can be obtained if interferometry is considered as a frequency- when the acoustic-laser technique is paired with other modulated measurement technique, this proposed penetrating NDT techniques. Previous attempts have photophone-based technique has the potential to be Cheng and Lau 143

modification, a photophone-based measurement device is conceived for analyzing surface mechanical vibration at a distance. Various parameters and their influence on SNR are investigated. The FRP-bonded specimen is excited acoustically and its frequency response is measured remotely using the aforementioned photophone- based device. The result corresponds well with theoretical prediction, allowing defect dimensions to be inferred based on the measured frequency response. If interferometry is considered as a frequency-modulated technique, this proposed technique has the potential to be developed into the amplitude-modulated counterpart of interferometry.

Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding The author(s) disclosed receipt of the following financial sup- port for the research, authorship, and/or publication of this article: This work was financially supported by the Croucher Foundation through the Start-up Allowance for Croucher Scholars with the Grant No. 9500012 and the Research Grants Council (RGC) in Hong Kong through the Early Career Scheme (ECS) with the Grant No. 139113.

References 1. Wu HF and Wan Y. New vision of the post-NIST civil infrastructure program: the challenges of next-generation construction materials and processes. In: Proceedings of the SPIE smart structures/NDE, San Diego, CA, 8–12 March 2014. 2. Bu¨yu¨ko¨ztu¨rk O, Gunes O and Karaca E. Progress on understanding debonding problems in reinforced con- Figure 9. (a) Top view of the setup of real-time measurement crete and steel members strengthened using FRP compo- 3 of a 50 3 50 3 30 mm defect using a smartphone, with the sites. Constr Build Mater 2004; 18(1): 9–19. light detector plugged in via a microphone port (white cable). 3. Bu¨yu¨ko¨ztu¨rk O and Yu T-Y. Understanding and assess- (b) The cyan curve is the real-time measurement and the yellow ment of debonding failures in FRP–concrete systems. In: curve is an averaged result. The observed peaks agreed with the Proceedings of the seventh international congress on advances previous results (Figure 5(a)). Inset: A copy of Figure 5(a) for in civil engineering, Istanbul, 11–13 October 2006. comparison. 4. Kotynia R. Debonding failures of RC beams strengthened with externally bonded strips. In: Proceedings of the developed into the amplitude-modulated counterpart international symposium on bond behavior of FRP in struc- of interferometry. tures, Hong Kong, 7–9 December 2005. 5. Smith ST and Teng JG. FRP-strengthened RC beams. I: review of debonding strength models. Eng Struct 2002; Conclusion 24(4): 385–395. 6. Teng JG and Chen J. Debonding failures of RC beams A low-cost augmentation to the acoustic-laser tech- strengthened with externally bonded FRP reinforcement: nique is developed and demonstrated in this study by behaviour and modelling. In: Proceedings of the first applying the technique for the detection of interfacial Asia-Pacific conference on FRP in structures (APFIS defects in FRP-bonded systems. As part of the 2007), Hong Kong, 12–14 December 2007. 144 Structural Health Monitoring 17(2)

7. Feng MQ, De Flaviis F and Kim YJ. Use of microwaves In: Proceedings of the 10th International Workshop on for damage detection of fiber reinforced polymer-wrapped Structural Health Monitoring (IWSHM), Palo Alto, CA, concrete structures. JEngMech2002; 128(2): 172–183. USA, 1–3 September 2015. 8. Maierhofer C, Brink A, Ro¨llig M, et al. Detection of shal- 21. Cheng TK and Lau D. Vibrational characteristics of low voids in concrete structures with impulse thermogra- FRP-bonded concrete interfacial defects in a low fre- phy and radar. NDT&E Int 2003; 36(4): 257–263. quency regime. In: Proceedings of the SPIE smart struc- 9. Maierhofer C, Krause M and Wiggenhauser H. Non- tures/NDE, San Diego, CA, 8–12 March 2014. destructive investigation of sluices using radar and ultra- 22. Lau D and Qiu Q. A review of nondestructive testing sonic impulse echo. NDT&E Int 1998; 31(6): 421–427. approach using mechanical electromagnetic waves. In: 10. Owen RD. Portable linear accelerators for X-ray and Proceedings of the SPIE smart structures/NDE, Las electron-beam applications in civil engineering. NDT&E Vegas, NV, 20–24 March 2016. Int 1998; 31(6): 401–409. 23. Qiu Q and Lau D. The sensitivity of acoustic-laser tech- 11. Pugliesi R and Andrade MLG. Study of cracking in con- nique for detecting the defects in CFRP-bonded concrete crete by neutron radiography. Appl Radiat Isotopes 1997; systems. J Nondestruct Eval 2016; 35: 33. 48(3): 339–344. 24. Yu T-Y, Cheng TK, Zhou A, et al. Remote defect detec- 12. Shaw M, Millard S, Molyneaux T, et al. Location of steel tion of FRP-bonded concrete system using acoustic-laser reinforcement in concrete using ground penetrating radar and imaging radar techniques. Constr Build Mater 2016; and neural networks. NDT&E Int 2005; 38(3): 203–212. 109: 146–155. 13. McCann DM and Forde MC. Review of NDT methods 25. Bruce RV. Bell: Alexander Graham Bell and the conquest in the assessment of concrete and masonry structures. of solitude. 1st ed. Boston, FL: Little, 1973. NDT&E Int 2001; 34(2): 71–84. 26. Bell AG. On the production and reproduction of sound 14. Akuthota B, Hughes D, Zoughi R, et al. Near-field by light. Am J Sci 1880; 118: 305–324. microwave detection of disband in carbon fiber reinforced 27. Bell AG. Selenium and the photophone. Nature 1880; polymer composites used for strengthening cement-based 22(569): 500–503. structures and disbond repair verification. J Mater Civil 28. Case TW. Infra red telegraphy and telephony. J Opt Soc Eng 2004; 16(6): 540–546. Am 1922; 6(4): 398–406. 15. Li J and Liu C. Noncontact detection of air voids under 29. Huxford WS and Platt JR. Survey of near infra-red com- glass epoxy jackets using a microwave system. Subsurf munication systems. J Opt Soc Am 1948; 38(3): 253. Sens Technol Appl 2001; 2(4): 411–423. 30. Yu T-Y and Haupt R. Damage inspection of fiber rein- 16. Rhim HC and Bu¨yu¨ko¨ztu¨rk O. Wideband microwave forced polymer-concrete systems using a distant acoustic- imaging of concrete for nondestructive testing. J Struct laser NDE technique. In: Proceedings of the SPIE smart Eng 2000; 126(12): 1451–1457. structures and materials nondestructive evaluation and 17. Popovics JS and Rose JL. A survey of developments in health monitoring, San Diego, CA, 7 March 2010. ultrasonic NDE of concrete. IEEE T Ultrason Ferr 1994; 31. Leissa AW. Vibration of plates. Washington, DC: Scien- 41(1): 140–143. tific and Technical Information Division, National Aero- 18. Clark MR, McCann DM and Forde MC. Application of nautics and Space Administration, 1969. infrared thermography to the non-destructive testing of 32. Chen JG, Wadhwa N, Cha YJ, et al. Modal identification concrete and masonry bridges. NDT&E Int 2003; 36(4): of simple structures with high-speed video using motion 265–275. magnification. J Sound Vib 2015; 345: 58–71. 19. Ham S and Popovics JS. Application of contactless ultra- 33. Wu HY, Rubinstein M, Shih E, et al. Eulerian video mag- sound toward automated inspection of concrete struc- nification for revealing subtle changes in the world. ACM tures. Automat Constr 2015; 58: 155–164. T Graphic 2012; 65: 1–8. 20. Cha YJ, Chen JG and Bu¨yu¨ko¨ztu¨rk O. Motion magnifi- 34. Wadhwa N, Rubinstein M, Durand F, et al. Phase-based cation based damage detection using high speed video. video motion processing. ACM T Graphic 2013; 80: 1–9.