Influence of Newtonian and non-Newtonian fluid behaviour on void fraction and bubble size for a gas-liquid flow of sub-millimeter bubbles at low void fractions

Abstract

This study compares the effect of Newtonian and non-Newtonian blood-mimicking test liquids on void fraction and bubble size distribution (BSD) for the case of gas–liquid bubbly flow. Along with this comparison, the novelty of the present work is extended to the examined conditions, which combine low void fractions (<10−1) with bubbles smaller than 1 mm, resembling bubbly flow in human bloodstream during Decompression Sickness. Such conditions, however, can be also found in common two-phase applications such as subcooled flow boiling in macro-channels. Experiments are performed in co-current upward bubbly flow. Void fraction and BSD are determined by a non-intrusive electrical impedance technique and by image analysis of bubbly flow images, respectively. Xanthan gum is employed (0, 150 and 1000 ppm) to impart non-Newtonian (shear-thinning) behaviour to the baseline aqueous glycerol/NaCl test liquid. The role of surface tension (by adding Triton X-100) and gas/liquid superficial velocities (Usg, Usl) on acquired data is studied as well. Results demonstrate that increase of bubble size and decrease of void fraction is primarily due to the non-Newtonian liquid character rather than the increase of dynamic viscosity caused by the addition of high Xanthan gum concentration. The addition of surfactant (Triton X-100) decreases bubble size and increases void fraction. Furthermore, bubble size and void fraction increase with Usg and decrease with Usl. Void fraction signal fluctuations show a strong dependence on bubble size. Furthermore, the performance of an empirical equation that predicts average bubble diameter from void fraction statistical quantities is tested and seems to be unaffected by the Newtonian/non-Newtonian liquid behaviour. Of particular interest are the intense peaks noticed in void fraction signals for the highest examined Xanthan gum concentration (1000 ppm) and the lowest examined liquid superficial velocity (2.89 cm/s). These peaks correspond to voluminous bubbles clusters rising among isolated bubbles, which do not appear at higher liquid velocities.