Ultra-highly electrically sensitive UHPC for defect detection: Electrical signal response and imaging

Abstract

Image visualization mainly focuses on structural surface damage monitoring, whereas detecting internal defects necessitate additional parameters. Therefore, this study proposes a detection method that combines electrode arrays and reconstruction algorithm for probabilistic inspection of damage (RAPID) to detect image defects in ultra-highly electrically sensitive UHPC (UHES-UHPC). UHES-UHPC, as a self-sensing concrete sensors intrinsically integrate sensing and data acquisition capabilities. The resistivity distribution curves and sensing pathways of UHES-UHPC were further analyzed with different defect locations and sizes. The RAPID algorithm and elliptical model were implemented for defect visualization, which improves the positioning accuracy and enhances the testing efficiency by fewer sensors than conventional ERT method. The full-summation imaging method was applied to eliminate the phenomenon of uneven probability distribution of the array (UPDA), enhancing the quality of imaging outcomes. Receiver operating characteristic (ROC) curves and probability value curves were used to objectively evaluate the performance of this method. The results showed the defects at different locations can be detected based on the intersection of the maximum resistivity sensing pathways, whilst the maximum detected deviation was 1.13 cm. For defects of different diameters, the minimum detection threshold was 0.6 cm, further defining the detection range. In addition, the imaging noise area was reduced by 34.4 % by the full-sum imaging method. The probability value curve showed that the localization accuracy of the imaged region reached 96 %, and the area under the ROC curve reached 0.96, which was close to the ideal value of 1. This NDT method can detect accurate defection and visualize the internal defects, which provides a theoretical basis for building structure health monitoring.