Effect of void fraction on capacitance of sensor for void fraction measurement

Abstract

Capacitance sensor is a non-intrusive devices used for the void fraction measurement in two-phase flows. Capacitance sensor is formed by two electrodes placed at the outside walls of a dielectric pipe, through which a two-phase medium flows. The capacitance of the sensor depends on the geometric arrangement of electrodes and distribution of the electrical field, which depends on the electrical permittivity of the medium around the electrodes and in the pipe. For a two-phase gas–liquid​ flow, the effective permittivity of dielectric between the electrodes varies due to the variation in the percentage of gas and liquid in the pipe. In this paper, the variation of sensor capacitance were determined numerically for the annular-flow pattern. The capacitance was calculated by the integration of electric field distribution around two symmetrical concave electrodes in the form of cylinder fragments using the Gauss law. This shape of electrodes was chosen because the sensitivity of this type of sensor is the highest, and is the most often used in practice. The results of numerical simulations were compared with experimental results. The capacitance changes due to void fraction variation were determined experimentally by the method of frequency deviation. It was noticed that regardless of complex geometry of the electric field between two concave electrodes, the electric field lines within the pipe walls do not deviate much from the parallel distribution. From the simulations for various arc lengths of the electrodes resulted that the changes of relative capacitance of the sensor are almost the same for all arc lengths, although the absolute value of capacitance of the sensor depends on this arc length of the electrodes. The numerical results on electric field distribution for annular flow also indicated that the model of two parallel capacitances, representing the liquid and gas phases, usually assumed in various papers as the most consistent with experimental results, can be used for the representation of two separate phases in two-phase flows.