The volumetric ultrasound imaging strategy for quantification of the defects in concrete based on the optimized cumulative kurtosis method

Abstract

Currently, ultrasonic tomography is widely used for quantifying the status of damage/flaw in various materials including metal, concrete, and composite. It is capable of visualizing the internal damage or flaws by reconstructing the velocity of ultrasound. For most cases, the conventional ultrasonic tomography is adapted for slice-based investigations. Though 3D image can be generated by slice-based method, the involvement of multi-scan interpolation caused high computational cost and unexpected errors. Moreover, for 3D imaging, two essential factors are dominant: (i) reconstructing the velocity; and (ii) scheme of forming up the 3D image. To this end, this study proposed to develop a novel volumetric ultrasound imaging strategy with high accuracy. To obtain more accurate Time-of-Flight (TOF) for ultrasonic velocity, an improved method based on cumulative kurtosis is proposed. With the proposed method, the adverse influence of internal complexity within samples on ultrasound was reduced. The proposed novel volumetric ultrasound imaging strategy was verified numerically and experimentally. Finally, the performance of the proposed method has been evaluated by the comparison of three-dimensional imaging results with different inclusion. Additionally, the parametric study was conducted using path average velocity, voxel velocity, and image accuracy. The results show a positive correlation between the number of voxels and imaging accuracy. However, as the number of voxels increases, the errors introduced by the Simultaneous Algebraic Reconstruction Technique (SART) increase. The influential factors on the imaging accuracy were discussed, such as inclusion eccentricity, the relationship among the reduced volume imaging quantity, and accuracy in voxel inversion result. For further application, the recommendations were also provided.