Search

The effect of channel diameter on two-phase flow was investigated to identify the phenomena which distinguish microchannels from minichannels. Experiments were conducted with a mixture of nitrogen gas and water in circular channels of 530, 250, 100, and 50 μm diameter. The temperature, pressure and flow rates of the liquid and gas were measured and images of the flow patterns recorded. The two-phase flow was characterized by the flow regime map, void fraction, and frictional pressure drop. In the 530 and 250 μm channels, the two-phase flow characteristics were similar to those typically observed in minichannels of ∼ 1 mm diameter. In the 100 and 50 μm channels, the two-phase flow behaviour departed from that observed in minichannels––the occurrence of slug flow dominated, the void fraction–volumetric quality relationship departed from a linear Armand-type correlation, and mass flux no longer influenced the two-phase frictional multiplier. Clearly, the channel diameter has an effect on two-phase flow in the range of channel diameters investigated. A new slug flow model is also proposed to gain physical insight into the observed flow characteristics in microchannels. The model can predict the two-phase frictional pressure gradient for the 100 and 50 μm channels, if the actual void fraction data are used, substantiating the assumed flow mechanism.

An adiabatic experiment was conducted to investigate the effect of channel geometry on gas-liquid two-phase flow characteristics in horizontal microchannels. A water-nitrogen gas mixture was pumped through a 96 μm square microchannel and the resulting flow pattern, void fraction and frictional pressure drop data were compared with those previously reported by the authors for a 100 μm circular microchannel. The pressure drop data were best estimated using a separated-flow model and the void fraction increased non-linearly with volumetric quality, regardless of the channel shape. However, the flow maps exhibited transition boundaries that were shifted depending on the channel shape.

An experimental investigation has been carried out on two-phase flow characteristics in a 100 μm diameter circular tube. Two-phase flow patterns were determined by video recording the flow in the transparent capillary tube made of fused silica, in which de-ionized water and nitrogen gas were injected at superficial velocities of jG=0.1–60 m/s for gas, and jL=0.02–4 m/s for liquid. Time-averaged void fraction and two-phase friction pressure drop data were also obtained and analyzed. The flow patterns observed were intermittent and semi-annular flows, but a closer study of the liquid film structure revealed gas core flows with a smooth or ring-shaped film and a serpentine-like gas core surrounded by a deformed liquid film. Bubbly and churn flow patterns were not observed. A flow pattern map was developed based on the probability of appearance of each type of flow, and compared to the existing flow pattern maps obtained for ∼1-mm diameter channels. Void fraction remained low even at high gas flow rates, indicating large slip ratios and weak momentum coupling between the phases. The single-phase friction factor and two-phase friction multiplier data were shown to be in good agreement with the conventional correlations.

Wave-turbulence interaction was experimentally investigated in turbulent open-channel flows with a shear-free wavy surface using a photochromic dye activation technique. In the experiments conducted, two-dimensional waves of different amplitudes, wavelengths, and frequencies were superimposed on a liquid surface via a mechanical wave maker. The range of Reynolds numbers varied from 3900 to 5000 based on the hydraulic diameter, with the corresponding aspect ratio of the channel width to liquid depth varying from 7.5 to 5.Within the range of Reynolds numbers investigated, the results showed that the streamwise turbulence intensity increased in the bulk and interfacial regions in comparison to the undisturbed flow.Furthermore, video sequences of the flow visualization experiments clearly revealed that the spanwise motion of the liquid was significantly suppressed; the traces did not immediately deform in the spanwise direction but retained their shape with increasing wave amplitude and frequency as compared to smooth interface flows. This suggests that waves may have suppressed longitudinal vortices generated near the smooth interface. The suppression of the longitudinal vortices in wavy open-channel flows has been proposed as a mechanism responsible for the turbulence energy redistribution, different from that for smooth open-channel flows.

The field of microfluidics is developing rapidly with advances in MEMS (micro electro mechanical system) and μ TAS (micro total analysis system) technologies. In various devices, controlling the flow rate of liquid or gas accurately at micro or nanoliter volume levels is required. In this work, the gas-liquid two-phase flow patterns in a microchannel T-junction have been examined. The two-phase flow patterns were observed in the T-junction under two different gas and liquid injection conditions. The flow patterns of liquid and gas in the microchannel after the T-junction were also observed and classified into three types depending on the flow rate, and a two-phase flow pattern map was constructed. The mechanism of bubble breakup at a micro T-junction was also experimentally and analytically investigated.Copyright © 2003 by ASME

An adiabatic experiment was conducted to investigate the effect of channel geometry on gas-liquid two-phase flow characteristics in microchannels. A mixture of water and nitrogen gas was pumped through a 96 μm × 96 μm square microchannel and the flow pattern, void fraction and pressure drop data were obtained and compared with those previously obtained in a 100 μm circular microchannel. The frictional pressure drop was determined from the measured total pressure drop, and the two-phase flow pattern and void fraction were determined from image analysis of the video recordings. In the square channel, 136 runs were performed over a range of 0.09 ≤ jG,AVG ≤ 62 m/s for the average superficial gas velocity and 0.01 ≤ jL ≤ 4 m/s for the superficial liquid velocity. The frictional pressure drop data showed that the calculations based on a separated–flow model were best at estimating the frictional pressure drop for both microchannels. No particular effect of the channel shape was found for the two-phase frictional pressure drop. The void fraction-to-volumetric quality relationship was also found to be similar for both shapes of microchannels, exhibiting an exponential increase in void fraction with increasing volumetric quality. The empirical correlation that describes the void fraction-to-volumetric quality relationship for the square microchannel was developed earlier from the measured data for the circular microchannel. Observations of the recorded images indicated the two-phase flow patterns to be primarily intermittent with liquid and gas slugs. The liquid film surrounding the gas core displayed a smooth or ring-like structure. The probability of each interfacial structure occurring was examined in detail to develop a novel flow pattern map consisting of four regions named slug-ring flow, ring-slug flow, multiple flow and semiannular flow. Between the square and circular microchannels, the two-phase flow maps exhibited transition boundaries that were shifted depending on the channel shape. The region of ring-slug flow that appears in the circular microchannel collapsed in the square microchannel, possibly due to the suppression of the liquid-ring film in the corners of the square channel.Copyright © 2003 by ASME

This paper presents a review of the flow characteristics in microchannels and our own flow visualization work using a photochromic dye activation technique. The review reveals differences in flow behaviour between micro- and macro-channels, and the need to measure instantaneous velocity profiles to explain these differences. In the experiment, de-ionized water containing a photochromic dye was pumped through a square capillary tube with inner dimensions of 96 μm × 96 μm to yield a Reynolds number of 0.1. A pulsed ultraviolet laser beam was used to create dye traces in the flowing liquid, and images of the traces were recorded with a video camera. Friction factor in laminar flow was also determined from pressure drop measurements at several Reynolds numbers using tap water. The experimental results showed the flow in intermediate-sized microchannels at low Reynolds numbers to still conform to conventional fluid mechanics theory, possibly with slight deviation. The success in obtaining quantitative results with this approach holds promise for further studies of flows in microchannels with smaller diameters.