Experimental evaluation of conductive flow imaging using magnetic induction tomography

Abstract

Multi-phase flow imaging is a challenging topic in industrial process tomography. In this paper, we present a non-invasive imaging technique for the electrically conductive phase of a multi-phase flow problem. Magnetic induction tomography (MIT) is sensitive to the conductivity of the target, and as such has the potential to be used as an imaging technique to visualise the conductive components in a multi-phase flow application. A 16 channel MIT system is used for this study, among which eight excitation coils are supplied with a 15V peak, 13MHz sinusoidal signal in sequence from a signal generator, while the remaining eight coils are floated as receivers. The imaging region of this MIT system has an inner and outer diameter of 190mm and 200mm respectively. Static fluid distribution patterns are produced using several fluids with different conductivities and placed inside the imaging region to form conductivity phase contrasts. Experimental results show within our hardware and software capability, a conductivity contrast of 0.06S/m for an inclusion that occupies 8.69% of the imaging region can be imaged. An in-depth experimental evaluation of the system response towards various fluid measurements is shown for the first time, as are results for quasi-static fluid experiments showing that a non-homogenous flow of gas bubbles can be imaged in various conductive backgrounds. In sum, the analyses presented investigate the feasibility and capability of MIT for this application, while also reporting some of the first flow rig tests in this field.

Simultaneous reconstruction of permittivity and conductivity using multi-frequency admittance measurement in electrical capacitance tomography

“…Both MIT and ECT are contactless methods. More work is still needed to establish MIT as a robust low conductivity imaging method [10]. In [11]capacitive coupling between MIT coils was investigated and highlighted a challenge in the interpretation of MIT data if phase changes are present due to capacitive coupling.…”

Computer methods for non-invasive measurement and control of two-phase flows: a review study

“…The first Electrical Capacitance Tomography (ECT) systems [120,113,158,79], Electrical Resistance Tomography (ERT) systems [33,32,147,127] as well as the Magnetic Induction (MI) [90,119] allowed to obtain only the rough evaluation of the process state because the information encoded in measurement data represents merely the fragment of the process which additionally was approximated into the cross-section surface through the sensor (2D ECT) [60,163,78,128,135]. This kind of imaging, however still commonly applied in many industrial applications [50,66,67,104,75,65], occurs to be insufficient from the process control point of view.…”

Total Variation Regularization With Split Bregman-Based Method in Magnetic Induction Tomography Using Experimental Data

“…Although standard Tikhonov is widely used for many applications, recently a hybrid Tikhonov regularization was shown to produce better imaging results [21]. To provide a fair comparison with the proposed TV algorithm we chose the most advanced hybrid Tikhonov method based on combining Laplacian and Tikhonov based regularization terms [21]…”

“…where 1 is an Laplacian regularization term, 2 is an identity matrix, and and are the regularization factors for 1 and 2 , respectively. The hybrid Tikhonov method has good quality in challenging low conductivity MIT data [21]. Here we empirically selected the regularization parameters for low conductivity and high conductivity reconstruction and used the same parameters in all experimental studies.…”