4. Flow regimes and heat transfer study
4.4. Results and discussion – Flow regime analysis
4.4.1. Working fluid study
4.4.1.1. Water at low pressure and low heat flux
The heat fluxes tested in the case of water at low pressure, Pr = 0.0002, ranged from
q = 7.5 – 63 kW/m2. At this low reduced pressure, the bubble departure diameter, λc, was
almost half the tube diameter and vapour production rates are relatively high. Single, explosive bubble nucleation is evident in both the flow visualisation and temperature and pressure recordings, for all levels of heat flux. These explosive boiling events give rise to a pseudo-steady oscillatory state. In this scenario, the thermosyphon is still capable of transferring heat, however the boiling events may cause shock damage to the thermosyphon components due to the pressure pulsations. Flow instabilities of this nature have been described in the previous literature, as outlined in Chapter 2. However, full field high-speed videography has not been performed so the precise mechanisms governing this behaviour have been largely speculative.
Figure 4.5 - Water at low pressure and low heat flux (parameter values shown in figure)
The position of water at low pressure and low heat flux in relation to the flow regime map is denoted by “X” in Figure 4.5. The high level of confinement for water at the current experimental conditions, coupled with the high rate of vapour production situates this state in the oscillatory geyser regime of two-phase flow.
The following discussion will focus on comparing the high-speed images of Figure 4.6 to the temperature and pressure measurements, shown in Figure 4.7. Figure 4.7 highlights the points along the pressure trace which correspond to the high-speed images.
Co
jv* = G/√[gDρv(ρl- ρv)]
Geyser flow Slug/Plug
flow
Bubbly flow Churn flow Water q = 7.5 kW/m2 P
r= 0.0002
Figure 4.6 - Water thermosyphon, q = 7.5 kW/m2, Pr = 0.0002, Δt = 0.06 s 1 2 3 4 5 6 7 8 9 T8 g C o nde ns er Eva p o ra to r
Figure 4.7 - Temperature and pressure trace for water at low pressure Pr = 0.0002, q = 7.5 kW/m2
The recorded images for water at the lowest heat flux of q = 7.5 kW/m2, over a period of
3 seconds during a typical geyser event are shown in Figure 4.6, with the associated data in Figure 4.7. The images are separated by a time step of Δt = 0.06 s and start from bubble nucleation in Figure 4.6 (1), to vapour slug growth (2 – 4), with the plug reaching a peak height in the tube at Figure 4.6 (6), before dropping back to the evaporator (7 – 9).
Firstly, considering Figure 4.7, it is clear that prior to bubble nucleation the system pressure is steady and the wall superheat of the evaporator section is increasing steadily. The superheat in the lower evaporator section, T8, is initially ΔTSH = 10 °C and rising with
time. During this prolonged waiting period, the liquid pool remains stagnant, represented here in Figure 4.6 image (1). The heat supplied to the evaporator section is transferred to the liquid pool, and stored within it, by conduction and possibly some natural convection, however this is not visible in Figure 4.6 with the current set-up. The wall superheat continues to increase until bubble nucleation occurs, in this instance at the bottom section of the pool, as seen in Figure 4.6 (1). At this point the wall superheat begins to decrease due to evaporative cooling. The high vapour generation rate causes the Taylor-like bubble to grow to such an extent that it fills the tube. This can be
0.01 0.02 0.03 0.04 0.05 0.06 10 20 30 40 5 6 7 8 Pres sure [bar] T [ C] Time [s] 1 23 4 5 6 7 8 9 Bubble nucleation T8 Tcond Tsat
attributed to the fact that there is a relatively high level of confinement at these test conditions.
Because nucleation occurred well below the free surface of the liquid, a large volume of liquid is trapped above the vapour slug, as depicted in Figure 4.6 (2). The high rate of bubble growth then forces this plug of liquid from the evaporator section to the condenser section. Thus, in contrast to conventional thermosyphon behaviour reported in existing literature, heat from the evaporator section is transported to the condenser as both sensible and latent heat. The notable decrease in wall superheat indicates an improvement in the heat transfer during the bubble growth phase as the bulk liquid begins to move.
With regard to the internal pressure of the system, as the wall superheat is increasing (between time 5 s and 6 s, Figure 4.7) there is very little change in the pressure. Attempts were made to measure the pressure in the evaporator though it was deemed impractical as it drained working fluid and influenced the boiling dynamics. This being the case, the sole pressure transducer is located at the top of the condenser section of the thermosyphon. Here the internal pressure is influenced by the movement of liquid plugs within the tube. As the liquid plug is forced upwards to the condenser, caused by the boiling event in the evaporator, the vapour in the condenser is compressed, resulting in an increase in the internal pressure in this region.
Due to the disparity between the liquid and vapour densities at this operating pressure (Table 4.1), evaporation results in a high rate of vapour production. The high vapour growth rate forces the liquid plug towards, and ultimately into, the condenser section. As the liquid plug is forced deeper into the condenser, the vapour, and possibly a small amount of non-condensable gas, in the top region of the condenser is compressed. This is evident in Figure 4.7 which shows that the pressure spikes between Figure 4.6 (3) and Figure 4.6 (6) increases sharply by 0.02 bar (57 % increase over quiet-phase pressure). As the liquid plug is cooled and returns back to the evaporator, the pressure reduces and returns to the original state. Once both temperature and pressure resume their state before the geyser event, the liquid pool is refilled and the cycle begins again.
As the bubble is forced into the condenser (Figure 4.6, 2 – 6) the meniscus of the plug tip is flat whilst the base of the plug is convex indicating that the pressure in the lower vapour region is higher than that of the upper region. Assuming quasi-static thermodynamic equilibrium, this will result in an increase in the saturation temperature. Since the wall temperature has dropped due to the evaporative cooling effect and the saturation temperature is increasing, the subsequent wall superheat decreases. In contrast, because the condenser wall temperature is relatively constant, the increased saturation temperature will result in an increased subcooling in the condenser. At Figure 4.6 (6) the liquid plug has reached its peak height in the tube and the pressure has also reached its peak magnitude. At this point the force due to the pressure acting on the base of the plug is insufficient to overcome the downward (pressure, gravity) and wall adhesion forces, and the plug begins to descend back to the evaporator. With changes in the forces acting on the plug, the top and bottom meniscus change shape, with the upper region being concave and the lower becoming flat. This would suggest that the pressure at the tip is now higher which, with the aid of gravity, forces the subcooled plug back to the evaporator.
Large amounts of energy were released when large liquid plugs collided with the end of the tube, or when two columns collided within the tube. Any vapour that had been present in the space between the columns rapidly collapsed. The energy from this collapse was released as sound and some was also transferred to the ends of the liquid columns. A ticking or clapping noise was audible during all of the water tests, a result of the water hammer effect occurring within the tube during oscillations. Most collisions were also followed by jets of liquid being pushed from opposite ends of the columns. This chain of events is depicted in Figure 4.8, with each successive image separated by Δt = 4 ms.
Figure 4.8 (1) shows the approach of two liquid columns, their direction of motion indicated by the black arrows on the left. On the instance of Figure 4.8 (2) a smacking sound was heard. Following this collapse of the vapour, the energy was transferred to the ends of the liquid column where distinctive jets were formed, image (3). Some heat from
and a vapour bubble forms and begins to grow, Figure 4.8 (4). The liquid jets also grow outwards, propelled by the initial energy release of the vapour collapse, Figure 4.8 (5 – 6).
Figure 4.8 - Liquid jets following column collision. Water, Pr = 0.0002, q = 7.5 kW/m2