olycu!'bc,note
3.3.2 Experimental results on large scale PHP
Several experiments have been conducted to determine the cooling capabilities of the pulsating heat pipe.
First the results of flow visualization are given. A comparison is made between the flow behaviour in the experiment and the model.
3.3.2.1 Flow visualization: comparison experiment and model
When camparing the flow behaviour in experiment and model the difference is clear. In the model a steady oscillating flow is obtained, negligibly influenced by orientation, whereas in the experiments this is not the case.
In the model in both horizontal and vertical orientation a steady oscillating flow is obtained.
When looking at the experiments the influence of gravity is important. In vertical orientation the flow is rather disordered. In horizontal orientation small oscillations around a fixed position occur when reaching an evaparator temperature of around 90 °C.
From video images the oscillation frequency is estimated to be about 10 Hz, which is about half ofthe frequency calculated by the model. The amplitude is about 1 cm.
In the experiments it is seen that when the pulsating heat pipe is in vertical position the liquid film on the walls drips down and is collected in the bottorn turn. In this way new liquid slugs are formed constantly, which affect the fluid flow.
In the model such phenomena do not occur. The model is strictly one dimensional, and no liquid film is modelled. Hence no new liquids slugs are formed, which divide a vapour plug into separate parts.
It is also seen in experiments that vapour plugs can coalesce, which decreases the amount of vapour plugs. In the model the liquid slugs are isolated from each other by the vapour plugs located between them. But in practice two vapour plugs can move towards each other to form one new vapour plug. The liquid slug, which was positioned between the plugs, is thus "squeezed away" through the liquid film, as is indicated in figure 3-14. This effect is also seen by Khandekar [26].
This may also affect the ratio of the forces acting on a liquid slug. In the model a pressure difference between the adjoining vapour plugscan only be levelled by a movement ofthe complete liquid slug. In practice the vapour plug at higher pressure may also squeeze liquid away as described above.
>
"Î I
I
i
I I
u
!I\
("'
~
I
Figure 3-14 Squeezing effect in the experiment
/!\
Iy+~ ,
I I
i I
3.3.2.2 Effect ofpulsating motion on heat throughput
The thermal resistance of the PHP is measured as a function of the input power. 1t has been seen that with increasing power and temperature difference between hot and cold si de, the thermal resistance drops [ 19].
In tigure 3-15 the evaporator temperature is given as a function of input power. An empty PHP is compared to a PHP in horizontal and vertical orientation with tilling ratio
(abbreviation: FR) of 20% and ethanol as working fluid.
The filling ratio FR is defined as
FR =-'q-·100%
~-~ota/ (3.29)
where Vliq is the volume occupied by the liquid phase and Ytotal the total volume of the pulsating heat pipe.
Almost no difference can be seen for the temperature rise of the evaporator for an empty PHP and a PHP in horizontal orientation. The small oscillations around a fixed position, as noted in the previous section, do not contribute to the heat transfer.
However when put in vertical orientation with the heater at the bottom, the evaporator temperature rises slow er at a power input of about 1 0 W and an evaporator temperature of 70
oe.
This corresponds to the temperature where a lot of disordered movement of the fluid is seen. At lower temperature no movement or only slow translation of vapour plugs 1s seen.The effect ofthe pulsating motion ofthe fluid can clearly beseen from tigure 3-15. In empty state the temperature reaches 100
oe
at a power of about 16 W. In verticalorientation 25 W can be dissipated without reaching 100
oe.
This is a power increase of over 50%.120
100
80
û
~ 60>
1-CD
40
20
0
,
.
..
0
,..
111 (\
"
"5 10
,.. •
,..
""
11
15 Power [Watt]
11
20
•
11•empty
• vertical; FR=20% 1 horizontal; FR=20% I
25 30
Figure 3-15 Mean evaparator temperafure as a function of dissipated power.
3.3.2.3 Fluid type
The thermal resistance is measured as a function of input power for different fluids.
The fluids which were used in the model are now used in the experiments. The results can be seen in tigure 3-16.
Figure 3-16 Thermal resistance as a function of input power for various fluids;
FR=45%, vertical orientation.
From the results can be seen that using methanol as working fluid gives the lowest thermal resistance. At an evaparator temperature of about 50
oe
movement of the working fluid could already be seen. Movement of water as working fluid was observed at a much higher evaparator temperature of 85oe.
For ethanol this value lies in between at 65oe.
The exact temperature at which fluid movement occurs is not predicted by the model (see section 3.2.2), however the trend is equal.
With methanol as working fluid a power input of 24 W results in an evaparator
temperature of 85°e (Rth
=
2.4 KIW), while only 16 W of power results in an evaparator temperature of 90°e for water.The ethanol filled flat plate pulsating heat pipe of Khandekar [20] had a thermal
resistance in vertical orientation which was 40 % of the unfilled heat pipe. In the present experiments the ethanol filled heat pipe has a minimal thermal resistance of 60 % of the unfilled equivalent.
This can be explained by the fact that in the present experiments the maximal evaparator temperature was about 90
oe,
while Khandekar measured up to 120oe.
The difference can be contributed to the fact that the thermal resistance drops with increasing power input.The aluminium plate with a filled PHP structure thus conducts or spreads heat up to 40 % better than an unfilled PHP. However about 40 % of the aluminium plate material is removed to create the PHP structure. When camparing the thermal resistance of the filled PHP to a bare aluminium plate the gain will be even lower.
3.3.2.4 Inclination angle
The inclination angle for a pulsating heat pipe can be of importance. The inclination angle ~ is defmed as depicted in figure 3-17.
Figure
When the evaporator region is positioned below the condensor, gravity will act on the liquid plug and liquid film on the tube wall. In this way liquid is being transported back to the evaporator by gravity, as is the case with thermosyphons.
In the case of horizontal position of the heat pipe gravity does not assist to bring liquid back to the evaporator region. Liquid transport then totally depends on the pulsating action of the fluid. In tigure 3-18 the influence of gravity can be seen on the heat cooling capability of the pulsating heat pipe. Experiments have been conducted to a power where the evaporator region reached a temperature near 1 00
oe.
7
Figure 3-18 Thermal resistance as a fonction of power at various inclination angles (FR=20%; ethanol)
Figure 3-18 shows the thermal resistance as a function of power and inclination angle. As can beseen the thermal resistance is almost constant for angles varying from 30 degrees to vertical. It was seen in experiments that only for angles close to horizontal ( <1 0) the thermal performance was affected. Forthese really smallangles the pulsating action diminishes to very small oscillations around a fixed position, rather than the disordered behaviour in non-horizontal orientations.
It was also seen that when put in a horizontal orientation evaporator dry-out could occur, meaning that there is no liquid left in the evaporator section. When tilted slightly after dry-out some liquid would enter the evaporator region and nucleate boiling effects could be seen. This was foliowed by vigorous movement of the fluid, resulting in a dramatic decrease of evaporator temperature from 95 to 70
oe.
The momentary thermal resistance is then even smaller than when operating in vertical conditions. After a certain period the vigorous movement would cease and normal operation is continued.3.3.2.5 Filling ratio
The filling ratio can affect the thermal performance. In tigure 3-19 the results can be seen for different filling ratios. The filling ratio is visually determined.
6
Figure 3-19 Thermal resistance as a fimction of input power at different filling ratios (Ethanol; vertical orientation)
The effect of filling ratio seems to be rather small. The values of 45 en 70 % FR are within 13 % of the thermal resistance values with a FR of 20 %. At power inputs above 10 W higher filling ratio decreases the thermal performance slightly. However filling ratio does not seem to appear a critica! parameter for this geometry, contrary to the case with standard heat pipes. This makes the filling process less critica!.