Simultaneous application of two laser-induced fluorescence approaches for film thickness measurements in annular gas-liquid flows

Abstract

This paper is devoted to the simultaneous application of two spatiotemporally resolved optical techniques capable of liquid film thickness measurements, namely Planar Laser-Induced Fluorescence (PLIF) and Brightness-Based Laser-Induced Fluorescence (BBLIF), to co-current downward annular gas-liquid flows. A single laser sheet is used to excite the liquid film, which has been seeded with a fluorescent dye, along a longitudinal/vertical plane normal to the pipe wall. Two cameras, one for each technique, are placed at different angles to the plane of the laser sheet in order to recover, independently by the two techniques, the shape of the gas-liquid interface along this section. The effect of the angle between the laser sheet and the PLIF camera axis is also investigated. In film regions where the gas-liquid interface is smooth and flat, the conventional approach used for interpreting PLIF data is affected by total internal reflection of the fluorescent light at the free surface, or “mirror effect”, which leads to an overestimation of the film thickness that increases as the angle between the laser sheet and the camera axis is decreased. Nonetheless, local features such as light intensity maxima or minima are often located within the fluorescent signals that correctly identify the true interface, which in these conditions also coincides well with the BBLIF film-thickness measurement. When a correction for the mirror effect based on simple flat-film optical calculations is applied, this leads to PLIF results that correspond well to the true film thickness. Interestingly, it is further found that interfacial three-dimensionality, and in particular azimuthal/circumferential non-uniformity, can lead to underestimation of film thickness by PLIF that in some cases counteracts the overestimation due to the mirror effect. Smaller angles between the laser sheet and camera axis make PLIF less susceptible to this error. In regions where the film surface is rough, including on the surface of disturbance waves, the mirror effect is suppressed. The BBLIF measurement, on the other hand, is affected by loss of signal sensitivity or saturation in thick film regions, as well as distortions at complex or multiple interfaces in agitated flow regions with significant wave activity. Direct comparisons between PLIF and BBLIF measurement data confirm that local overestimations of the film thickness by the latter occur in film regions with higher interfacial curvature, especially at the front slopes of waves and around gas bubbles entrained in the liquid even in smooth films, while underestimations occur in regions with multiple interfaces and inside bubbles although it is often the case that features such as bubbles can be identified and corrected for in simple flows with smooth interfaces. Correction procedures are developed to compensate distortions caused by both methods that make these techniques more accurate for standalone employment. In highly complex, three-dimensional flows, the simultaneous application of both techniques is highly recommended to attain the most reliable information.