The main topics of this dissertation are the hardware and the software developments in magnetic induction tomography imaging techniques. In the hardware sections, all the tomography systems developed by the author will be presented and discussed in detail. The developed systems can be divided into two categories, according to the property of the target imaging materials: high conductivity materials and low conductivity materials. Each system has its own suitable application, and each will thus be tested under different circumstances. In terms of the software development, the forward and inverse problems have been studied, including the eddy current problem modeling, sensitivity map formulae derivation and iterative/non-iterative inverse solvers equations. The Biot-Savart Theory was implemented in the ‘two-potential’ method that was used in the eddy current model in order to improve the system’s flexibility. Many different magnetic induction tomography schemes are proposed for the first time in this field of research, their aim being to improve the spatial and temporal resolution of the final reconstructed images. These novel schemes usually involve some modifications of the system hardware and forward/inverse calculations. For example, the rotational scheme can improve the ill-posedness and edge detectability of the system; the volumetric scheme can provide extra spatial resolution in the axial direction; and the temporal scheme can improve the temporal resolution by using the correlation between the consecutive datasets. Volumetric imaging requires an intensive amount of extra computational resources. To overcome the issue of memory constraints when solving large-scale inverse problems, a matrix-free method was proposed, also for the first time in magnetic induction tomography. All the proposed algorithms are verified by the experimental data obtained from suitable tomography systems developed by the author. Although magnetic induction tomography is a new imaging technique, it is believed that the technique is well developed for real-life applications. Several potential applications for magnetic induction tomography are suggested. The initial proof-of-concept study for a challenging low conductivity two-phase flow imaging process is provided. In this thesis, a range of contributions have been made in the field of magnetic induction tomography, which will help the magnetic induction tomography research to be carried on further.

Electrical capacitance tomography (ECT) is a technology that images the dielectric permittivity distribution of materials under test. ECT has been used as a tool for process monitoring in particular for two-phase flow measurement. These applications mainly focus on the dielectric samples, whose conductivity is negligibly small. This thesis studies ECT imaging with conductivity considerations. The conductive materials will affect the capacitance measurements and introduce difficulties in the ECT image reconstruction. This thesis presents solutions based on ECT to image material of different values of conductivity in different practical process or monitoring scenarios: the conductivity within materials under test is considered to be higher than 10^6 S/m, or less than 10 S/m. This work consists of the following innovative steps. (i) Through an ECT monitoring, floating (i.e., electrically non-grounded) metallic samples are imaged as dielectric illusions and the analysis of capacitance measurements over the conductors is delivered. (ii) Magnetic induction tomography (MIT) is firstly used for locating grounded metallic samples, thereafter as an assistant method to guide ECT to image the dielectric components. (iii) In low conductivity case MIT, as an indicator of conductive material again, helps ECT to solve multiphase flow problems. (iv) The multi-frequency complex ECT measurement provides a potential method to improve the ECT imaging ability for both conductive and dielectric materials. The first three ideas have been testified by both simulated and experimental results, while the fourth part is simulation-based results only on current stage.

Magnetic induction tomography (MIT) is a non-invasive, soft field imaging modality that has the potential to map the electrical conductivity (σ) distribution inside an object under investigation. In MIT, a number of exciter and receiver coils are distributed around the periphery of the object. A primary magnetic field is emitted by each exciter, and interacts with the object. This induces eddy currents in the object, which in turn create a secondary field. This latter is coupled to the receiver coils and voltages are induced. An image reconstruction algorithm is then used to infer the conductivity map of the object. In this thesis, the application of MIT for volumetric imaging of objects with low conductivity materials (< 5 Sm-1) and dimensions < 1 m is investigated. In particular, two low conductivity applications are approached: imaging cerebral stroke and imaging the saline water in multiphase flows. In low conductivity applications, the measured signals are small and the spatial sensitivity is critically compromised making the associated inverse problem severely non-linear and ill-posed.The main contribution from this study is to investigate three non-linear optimisation techniques for solving the MIT inverse problem. The first two methods, namely regularised Levenberg Marquardt method and trust region Powell's Dog Leg method, employ damping and trust region strategies respectively. The third method is a modification of the Gauss Newton method and utilises a damping regularisation technique. An optimisation in the convergence and stability of the inverse solution was observed with these methods compared to standard Gauss Newton method. For such non linear treatment, re-evaluation of the forward problem is also required. The forward problem is solved numerically using the impedance method and a weakly coupled field approximation is employed to reduce the computation time and memory requirements. For treating the ill-posedness, different regularisation methods are investigated. Results show that the subspace regularisation technique is suitable for absolute imaging of the stroke in a real head model with synthetic data. Tikhonov based smoothing and edge preserving regularisation methods also produced successful results from simulations of oil/water. However, in a practical setup, still large geometrical and positioning noise causes a major problem and only difference imaging was viable to achieve a reasonable reconstruction.

Made available in DSpace on 2016-12-12T17:38:38Z (GMT). No. of bitstreams: 1 JULIA GRASIELA BUSARELLO WOLFF.pdf: 4663430 bytes, checksum: 6c78e4867094b8623e5a0412abf9b544 (MD5) Previous issue date: 2011-08-02 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Magnetic induction tomography is a developing technique for visualization of electromagnetic properties of an object. It differs from others imaging techniques from electromagnetic signals because it is a noninvasive and contactless method whose main purpose is to obtain the distribution of conductivity in objects, which has important applications in industry and medicine. This master s degree dissertation discusses the development of the impedance method in three dimensions and its use in a modeling system of magnetic induction tomography. This method allows the calculation of induced currents in an object from an applied magnetic field, using valid approximations for quasi-static conditions of operation, which allows the modeling of the medium by a three-dimensional network of impedances. Thus, the impedance method was used for the calculation of eddy currents while the primary and secondary fields were calculated using the Biot-Savart Law. A simulation program in the C language has been developed to obtain the current and field distributions and calculate the sensitivity matrix of the system. The simulation has been compared with analytical models and experimental results. Two methods have been used to obtain images of the conductivity distribution from the simulated values of phase variation in signals from sensors in the tomograph: Tikhonov regularization and an original method proposed in this study called successive approximations. The methods are discussed and their results compared. It has been found that both underestimate the conductivity value, but they provide images with good linear correlation with respect to the real object. The successive approximations method has slightly better performance. / A Tomografia de Indução Magnética é uma técnica em desenvolvimento para a visualização das propriedades eletromagnéticas de um objeto. Difere de outras técnicas de produção de imagens a partir de sinais eletromagnéticos por se tratar de um método não invasivo e sem contato cuja sua finalidade principal é a obtenção da distribuição de condutividade em objetos, o que encontra aplicações importantes na indústria e na medicina. Essa dissertação aborda o desenvolvimento do método das impedâncias em três dimensões e o seu uso na modelagem de um sistema de tomografia de indução magnética. Este método permite o cálculo de correntes induzidas em um objeto a partir de um campo magnético aplicado, usando aproximações válidas para as condições quaseestáticas de operação, o que permite a modelagem do meio por uma rede tridimensional de impedâncias. Assim, o método das impedâncias foi utilizado para o cálculo das correntes eddy enquanto que os campos primário e secundário foram calculados usando a lei de Biot-Savart. Um programa em linguagem C foi desenvolvido para realizar as simulações, obter as distribuições de corrente e campos e calcular a matriz de sensibilidade do sistema. Os resultados das simulações foram confrontados com modelos analíticos e com resultados experimentais. Dois métodos foram usados para obter imagens da distribuição de condutividade a partir de valores simulados da variação de fase nos sinais obtidos nos sensores do tomógrafo: a regularização de Tikhonov e um método original proposto neste estudo denominado aproximações sucessivas. Os métodos são discutidos e seus resultados comparados. Verificou-se que ambos subestimam o valor da condutividade, mas fornecem imagens com boa correlação linear em relação ao objeto real. O método de aproximações sucessivas apresenta desempenho levemente superior.

Fiber-reinforced plastics for industrial applications face constantly increasing demands regarding efficiency, reliability, and economy. Furthermore, it was shown that fiber-reinforced plastics with tailored reinforcements are superior to metallic or monolithic materials. However, a trustworthy description of the load-specific failure behavior and damage evolution of composite structures can hardly be given, because these processes are very complex and are still not entirely understood. Among other things, several research groups have shown that material damages like fiber fracture, delamination, matrix cracking, or flaws can be discovered by analyzing the electrical properties of conductive composites, for example, carbon fiber–reinforced plastics. Furthermore, it was shown that this method could be used for structural health monitoring or nondestructive evaluation. Within this study, magnetic induction tomography, which is a new imaging approach, is introduced in the topic of nondestructive evaluation of carbon fiber–reinforced plastics. This non-contacting imaging method gains the inner spatial distribution of conductivity of a specimen and depicts material inhomogeneity, like damages, not only in two-dimensional images but also in three-dimensional images. Numerical and experimental investigations are presented, which give a first impression of the performance of this technique. It is demonstrated that magnetic induction tomography is a promising approach for nondestructive evaluation. Potentially, it can be used for fabrication quality control of conductive fiber–reinforced plastics and as a structural health monitoring system using an integrated or superficially applied magnetic induction tomography setup.