The influence of permittivity models on phantom images obtained from electrical capacitance tomography

Four different phantoms have been placed within the sensor field of an electrical capacitance tomography (ECT) system mounted on a bench-scale fluidized bed. The phantoms were created with fluid catalytic cracking catalyst particles having a packed-bed relative permittivity of 2.4. Linear back-projection (LBP) image reconstruction has been compared with iterative LBP for three permittivity models: parallel, series, and Maxwell model. The parallel permittivity model provides the best representation of the phantoms. Tubes of solids placed in an empty bed could only be reconstructed near the ECT wall, while voids created by immersion of an empty tube in a packed bed of solids could be imaged equally well at the center or wall. After 500 iterations, the size of the phantom in the reconstructed image is in very good agreement with the actual size of the phantom. (Author)