Electrical Resistance Tomography Imaging of Concrete

Cracking and deterioration of concrete are the leading causes of a premature failing of reinforced concrete structures. To assess the condition of concrete, a variety of destructive and non-destructive testing methods have been developed. From these two methods, the non-destructive testing (NDT) is a more favorable (albeit more challenging) option since the tested target is left undamaged. The NDT modalities include acoustic, electromagnetic and radiation based techniques. In this thesis, the feasibility of electrical resistance tomography (ERT) for NDT of concrete is studied. In ERT, electric currents are injected into the target through electrodes that are are attached to the boundary of the target. The resulting voltages between the electrodes are measured and this boundary voltage data is then used to reconstruct the internal conductivity distribution of the target. The reconstruction of the internal conductivity distribution of concrete is expected to provide valuable information about the condition of the structure so that appropriate repairs can be taken in time. The difficulty in ERT, as is in any other diffuse tomography modality, is that the problem has a nature of an ill-posed inverse problem. This implies that the solutions of the problem are unstable and nonunique in the classical sense. As a consequence, extra attention must be directed to the mathematical modeling of the measurements as well as to the reconstruction methods. Furthermore, concrete is strongly heterogeneous material composed of cement matrix, aggregate and different chemical compounds that create a challenging target for electrical modalities. Previous studies have shown that ERT is a potential tool for NDT of concrete, but the quality of the reconstruction was not yet sufficient for practical applications. In this thesis, especially localizing reinforcing bars and crack identification are considered. To meet those aims, novel computational methods for the image reconstruction are developed. For the crack identification and localizing reinforcing bars, a novel adaptive meshing approach was developed. In the new approach cracks and reinforcing bars are modeled as internal structures. The results show that by employing accurate mathematical models and statistical inversion techniques based on the Bayesian framework, ERT can become an applicable tool for practical NDT of concrete. Universal Decimal Classification: 537.311.6, 620.179.1, 620.19, 621.3.011.2, 621.317.33, 691.32 INSPEC Thesaurus: nondestructive testing; flaw detection; concrete; imaging; tomography; electric impedance imaging; electric resistance measurement; electrical conductivity; inverse problems; image reconstruction; Bayes methods Yleinen suomalainen asiasanasto: rikkomaton aineenkoetus; kuntokartoitus; betoni; betonirakenteet; kuvantaminen; tomografia; resistanssi; sähkönjohtavuus; impedanssitomografia; bayesilainen menetelmä