Novel three-dimensional ultrasonic-imaging-based concrete defect localisation and quantification technique considering concrete inhomogeneity

Three-dimensional (3D) ultrasonic imaging is a non-destructive testing method used to locate and quantify defects in concrete. To address the low-accuracy issue caused by imaging artefacts, this paper first considers the inhomogeneity of concrete and its uncertainty in the distribution of internal defects and then improves the algebraic reconstruction technique (ART) algorithm by correcting iterative initial value of ultrasonic velocity and the travel time error factor, thereby reducing the impact of noise artefacts. The image post-processing method of multi-attribute clustering of ultrasonic velocity and attenuation coefficient is then proposed to eliminate defect artefacts. Finally, the inverse distance weighting (IDW) algorithm is used to achieve 3D imaging. Single defect detection results show that the improved ART algorithm reduces the coefficient of variation value in the velocity of the defect region from 10.98 % to 2.35 %. Additionally, the image post-processing method, based on clustering ultrasonic velocity and attenuation coefficient, achieves image fitting degrees of 96.8 % and 96.9 %, respectively. The 3D ultrasonic imaging method, based on the IDW algorithm, achieves fitting degrees of 97.03 % and 95.57 %. Moreover, application of the proposed method in the steel reinforced concrete frame node further demonstrated that this method can locate and quantify internal defects for concrete structures.