Slug flow, characterized by elongated bubbles flowing in the axial
direction, and gas core flow, which is defined as long gas slug
occupied over a length of channel, were observed. As shown in
Fig. 7a, for the inlet section which led the gas–liquid mixture directly
to the micro-channels, large gas bubbles in the chamber
could expand along the channel length, leading to gas core flow
in most channels. The formation of gas core flow is schematically
depicted in Fig. 8. Comparatively, almost all the channels could
be occupied by slug flow, as seen in Fig. 7b, when a piece of foamed
plastic polymer inserted in the inlet section was used to induce the
discrete bubbles before entering the channels.
In this work, the flow pattern results were taken simultaneously
with the heat transfer data. Flow patterns were expected to affect
the heat transfer phenomena. Different flow patterns may give differences
in heat transfer mechanisms during gas–liquid flow. The
effect of flow pattern on heat transfer characteristics is discussed
next.
4.2. Heat transfer results
The average Nusselt numbers were plotted against superficial
Reynolds numbers, as presented in Figs. 9 and 10. As seen in
Fig. 9, the results, as expected, showed monotonous increases of
the Nusselt numbers with the liquid superficial Reynolds numbers.
As also shown in the figure, the Nusselt number increases with
increasing the gas superficial Reynolds number. Hetsroni et al.
[19] studied heat transfer characteristics in the channels of
0.1 mm in size and reported that an increase in gas superficial Reynolds
number involved a decrease in heat transfer, which was
opposite to the present results corresponding to micro-channels
of 0.4 mm in size. The behavior observed by Hetsroni et al. [19]
may be mainly due to, as suggested by Chung and Kawaji [2], significant
effects of surface tension as well as liquid viscosity, which
prohibit agitation of the gas–liquid interface in very small channels.
Such different observations were also reported by Choo and
Kim [21]. Based on their work, the heat transfer results for the
channels with diameters of 0.14 mm and 0.22 mm revealed the
trend similar to that of Hetsroni et al. [19]. On the other hand,
for the 0.33 mm and 0.5 mm channels, Choo and Kim [21] reported
the Nusselt number increasing with increasing the air flow rate.
The effect of flow pattern on gas–liquid heat transfer is shown
in Fig. 10. The Nusselt number was higher for slug flow (Fig. 7b)
than for gas core flow (Fig. 7a). This result indicates that the installation
of a piece of foamed plastic polymer can improve the distribution
of phases in micro-channels. As shown in Fig. 10, the
Nusselt number can be enhanced up to 40% at high flow rates.
The improvement of heat transfer may be mainly attributed to
the small gas slugs contributing to agitation in the liquid film on
the wall.
There have been several factors influencing the two-phase flow
maldistribution in parallel channels. The relevant factors were recently
given in details by Dario et al. [26].
Interestingly, the results based on micro-scale flow boiling
involving complex mechanisms, reported by Ong and Thome [27]
and Saisorn et al. [6], indicated that slug flow appeared with low
heat transfer coefficient in comparison to the vapor core flow.
Fig. 11 illustrates heat transfer enhancement ratio for the present
flow conditions. For slug flow, a heat transfer enhancement up
to 80% was obtained over the single-phase flow.