Electrical resistance tomography for mortar with multiple rebars structure investigation

Reinforced concrete has been broadly used as structural material in every part of the world such as in dams, nuclear reactor, stadium and pavements. Another type of concrete is mortar but without a coarse aggregate. The presence of reinforcement bars (rebars) in concrete or mortar play an important role in preventing the structure from breaking apart easily. Non-destructive testing (NDT) has been well known methods in inspect and monitor the condition of concrete and mortar structure due to no damage effect on the structure. NDT techniques consist of electromagnetic, radiography and electrical technique are capable of locating reinforcement bars. However, electrical technique is the most suitable since this technique is non-hazardous, non-invasive, and fast response in comparison to electromagnetic and radiography technique. Previous research has shown that electrical methods such as tomographic modalities using electrical resistance tomography (ERT) have yet to be used to image multiple rebars for mortar structures. Therefore, in this thesis, ERT system is proposed to perform imaging of multiple rebars for mortar structure. The ERT belongs to soft-field tomography and the nature of soft-field tomography possesses ill-posed inverse problem that produce low spatial resolution image. Iterative image reconstruction algorithms, such as Total Variation (TV), have been proposed to solve this problem. Iterative image reconstruction is beneficial in offering high spatial resolution image. The TV algorithm was compared to four non-iterative algorithms, consist of Linear Back Projection (LBP), Laplace, Newton One-Step Error Reconstruction (NOSER), and Tikhonov, to prove that the TV algorithm is suitable for imaging multiple rebars and solving low spatial resolution. All the algorithms were applied and tested under four different profiles. All the images produced was analysed using image assessment tool such as Mean Structural Similarity (MSSIM). Global thresholding method was adapted into all the images to further enhance the visibility of the rebars. Coefficient of Correlation (CoC) method was introduced to assist the thresholding method to find the optimise thresholding value. In comparison to non-iterative algorithms, the TV algorithm has exhibited outstanding performance in the MSSIM index from the simulation and experimental studies. The use of global thresholding in conjunction with the CoC approach has shown to have a significant effect on eliminating unnecessary artefacts and the rebars visibility for all the algorithms. As compared to non-iterative algorithms, the TV algorithm has shown superiority in the MSSIM index even after the implementation of thresholding. The proposed TV algorithm has demonstrated the ability to successfully visualise the rebars better based on the MSSIM index trend developed from simulation and experimental studies as well as after thresholding deployment. Moreover, the TV algorithm has proven also to be a solution to overcome the nature of low spatial resolution in ERT imaging and capable to visualise the multiple rebars for mortar structure.