Influence of orientation upon the hydrodynamics of gas–liquid flow for square channels in monolith supports

Monoliths are being used increasingly as catalyst supports for two-phase gas–liquid reactions, yet substantial differences in the mass transfer performance between different configurations have not been thoroughly explained using either mass transfer or hydrodynamic arguments. In this paper, investigations of the differences in hydrodynamics between up-flow and down-flow have been made in a single channel using square glass capillaries of either 1.5 or 2 mm section. The fluids used were either water or 30%v/v30%v/v isopropanol/water mixture and air. Predictive flow maps are presented for down-flow: annular, Taylor (slug) flow, bubbly and churn flow were observed. In the Taylor flow regime, slug velocities and lengths measured using an optoelectronic technique were found to be in good agreement with the drift flux model [Zuber, N., Findlay, J.A., 1965. Average volumetric concentration in two-phase flow systems. Journal of Heat Transfer 87, 453–468]. Non-zero drift velocities were obtained. Particle image velocimetry (PIV) was used to investigate the velocity fields within the liquid slugs. For short slugs (slug length less than the tube hydraulic diameter), a flow is developed where the axial velocity component is only a function of position in the tube cross-section. The velocity profile is relatively flat, with the maximum observed velocity at the axis of the tube, VmaxVmax, being 0.8–1 times the bubble velocity, VBVB. For long slugs, the axial velocity component depends on both the axial position and the position in the tube cross-section. Close to parabolic profiles are developed with Vmax/VB≈1.1–1.7Vmax/VB≈1.1–1.7. The location of the centre of the recirculation vortices produced in long slugs was found to be closer to the tube centre in down-flow compared with up-flow. Recirculation times in up-flow were 3 times faster: this has implications for the models used to predict rates of mass transfer and residence time distribution.