ACI Materials Journal/May-June 2010 305 ACI MATERIALS JOURNAL TECHNICAL PAPER ACI Materials Journal, V. 107, No. 3, May-June 2010. MS No. M-2009-227 received June 29, 2009, and reviewed under Institute publication policies. Copyright © 2010, American Concrete Institute. All rights reserved, including the making of copies unless permission is obtained from the copyright proprietors. Pertinent discussion including authors’ closure, if any, will be published in the March-April 2011 ACI Materials Journal if the discussion is received by December 1, 2010. The feasibility of impact-generated Rayleigh waves (R-waves) for measuring deep surface-opening cracks in concrete structures was studied. The aim is to contribute to a methodology for simple and effective in-place crack depth estimation. Specimens induced with vertical slits of different depths were prepared for measurement. A two-sensor array was implemented and elastic waves of different central frequencies were generated by mechanical impacts with steel-ball hammers of different ball diameters. R-wave amplitudes were extracted from the waveforms. Attenuation of R-waves due to diffraction and scattering by the slits and the trend of amplitude decaying with slit depth were examined. A reasonable correlation between the amplitude factor and slit depth-to-wavelength ratio was established, which indicated the loss of sensitivity in the change of amplitude factor with regard to dominant wavelengths smaller than the slit depth. By comparing the results of P-wave time-of- flight (TOF) method, the results by measuring again using the proposed method confirmed the feasibility of R-wave attenuation as an alternative parameter for characterizing surface-opening cracks. In addition, it was also demonstrated potential problems associated with the reliability of P-wave TOF method in estimating a crack with limited length. Keywords: amplitude factor; attenuation; nondestructive testing; Rayleigh waves; surface-opening cracks. INTRODUCTION Concrete structures are susceptible to various forms of attacks including applied load, drying shrinkage, temperature variation, steel corrosion and other forms of degradation. Surface cracking is one of the most common forms of defect found in concrete structures as a consequence of the attacks. Depending on the extent and their locations, surface-opening cracks pose different levels of threats in degrading the overall strength and durability performance of the structure. To make things worse, these cracks, if they penetrate through the thickness of the concrete cover, could result in exposure of steel reinforcement and facilitate the ingress of chloride to accelerate the corrosion process. 1 Effective detection and evaluation of surface-opening cracks are vital so that appropriate repair or strengthening can be conducted to restore structural integrity. Several nondestructive tests (NDT) have been developed to detect surface-opening cracks in concrete structures. The ultrasonic time-of-flight (TOF) method is the most straightforward to adopt, in which the time and distance of compression wave (P-wave) propagation are used to compute the velocity, so that anomalies that cause delay in the wave propagation can be identified. Van Hauwaert et. al. 2 conducted a study to follow cracks developed under bending of reinforced concrete (RC) members by establishing a direct relation between the variation of ultrasonic parameters during crack growth and the crack length. There are also reports on the development of ultrasonic spectroscopy for evaluating cracks and an ultrasonic imaging method that scans the concrete surface for cracking. 3-5 A similar method to the ultrasonic TOF method was developed by using elastic waves induced by mechanical impacts, in which the travel time of P-waves of known velocity diffracted from crack tip were measured to calculate the depth of crack. 6,7 In most real cases, however, the crack surfaces were filled with water and dust or in contact with the opposite surface, as well as in connection with steel reinforcements. The filling, contact, or connection can carry the P-waves across the cracks. Therefore, when identifying the depth of cracks, especially those with great depths, underestimation can commonly occur due to the shorter-than-expected wave path through the bridging points. 8 Recently, studies on the use of Rayleigh wave (R-wave) scattering and attenuation characteristics in estimating depths of surface-opening cracks in concrete have been reported. 9-16 R-waves propagate the surface of an object with a penetration depth of approximately one wavelength. Furthermore, R-waves carry a higher amount of energy than bulk waves, that is, approximately 67% of the total energy of elastic waves, and have a low attenuation rate in geometric spreading due to their cylindrical wavefront, which enable them to propagate over longer distances than the bulk waves. 9 Also, unlike P-waves, the cylindrical particle movements of R-waves, which are perpendicular to the propagation direction, depend heavily on the shear properties of the medium. 14 Therefore, the existence of steel bar, water, or dust, as well as contact between aggregates inside a crack, are not expected to affect the wave propagation because of insignificant shear properties of the substance and contact surface. Most of the studies for R-waves, experimentally or analytically, have been linked with evaluating concrete cracks that are relatively shallow in depth. In these studies, measurements by ultrasonic excitations of high frequencies were conducted. Although signal processing and identification for crack evaluation can be precisely performed with high consistency due to the good sensitivity of high frequency waves toward attenuation, penetration of wave energy into concrete has been very limited due to the short wavelength of the ultrasonic waves. It is also known that the relation between amplitude decay and crack depth can be established by taking the ratio between crack depth and wavelength; therefore, the relation is not confined to a certain range of crack depth, that is, varying the wavelength enables crack-depth evaluation of any scale. 9-13,17 There is still a lack of experimental data, however, to confirm the effectiveness of Title no. 107-M36 Characterization of Deep Surface-Opening Cracks in Concrete: Feasibility of Impact-Generated Rayleigh-Waves by Hwa Kian Chai, Shohei Momoki, Dimitrios G. Aggelis, and Tomoki Shiotani