3.1. The Effects of Power Distribution on the Threshold Critical Power of the Heat Pipe Systems
Figure 3 shows that the threshold critical power of the heat pipe systems was experimentally investigated under different power distributions, including heat load A and heat load B. In the experiment section, heat load B was introduced as an auxiliary heat source to make the threshold critical power of the heat pipe systems adjustable. Therefore, the additional heat load B is similar to a thermal switch. As shown in
Figure 3a,b, the surface temperature of two heat pipe systems increased steadily with heat load A. Additionally, there was little difference in the threshold critical power of them without heat load B. For example, the threshold critical power of TSHP was 55 W, while TGHP was 50 W. It could be known that the same length of heat pipe might have almost the same threshold critical power, even though they had different structures. For example, TSHP had a necking section with a length of 30 mm and a diameter of 4 mm. TGHP was only a long straight grooved tube without a necking section. There were some interesting things that occurred that the surface temperature difference of the heat pipes showed obvious distinguish. Nine points for measuring the temperature of TSHP had good consistency, which meant that its thermal performance was pretty good. However, TGHP had good temperature consistency under the low operating power, while it obtained a large temperature deviation under the high operating power.
Figure 3c,d demonstrated the threshold critical power of two heat pipe systems under different heat load B inputs in detail, respectively. Additionally, two kinds of the filling working liquid mass for two heat pipe systems were also experimentally investigated, respectively. In this study, the filling working liquid mass was not the key research factor. Therefore only two kinds of the filling working liquid mass for two heat pipe systems were chosen to show how it influenced the thermal performance of the heat pipe systems. Then, the heat pipes with better thermal performance were selected to push ahead with the following experiments. Two kinds of the filling working liquid mass (1.05 g and 0.95 g) for TSHP under different heat loads B were demonstrated as the following
Figure 3c. For convenience, TSHP, with the filling working liquid mass of 0.95 g, was named TSHP-1, while the other was named TSHP-2. As mentioned previously, heat load B played a significant role in the threshold critical power adjustment of the heat pipe system. TSHP showed almost the same change with heat load B. Their threshold critical power increased rapidly with low heat load B and increased slowly or even kept constant with high heat load B. For example, both TSHP-1 and TSHP-2 had a large growth of 10 W with heat load B input less than 4 W. It was almost a 20% increase in threshold critical power for TSHP-1 and TSHP-2 (from 55 W to 65 W for TSHP-1; from 50 W to 60 W for TSHP-2). When heat load B increased from 4 W to 12 W, both TSHP-1 and TSHP-2 increased 5 W (from 65 W to 70 W for TSHP-1; from 60 W to 65 W for TSHP-2). Moreover, TSHP-1 could be speculated better than TSHP-2 in the threshold critical power with heat load B increase. As a result, TSHP-1 was chosen to carry on the following experiments. Two kinds of the filling working liquid mass (1.5 g and 1.1 g) for TGHP under different heat loads B were demonstrated as the following
Figure 3d. For convenience, TGHP with the filling working liquid mass of 1.5 g was named TGHP-1, while the other was named TGHP-2. Their threshold critical power increased slowly or even kept constant with low heat load B and increased rapidly with high heat load B, which was in contrast with TSHPs. For example, both TGHP-1 and TGHP-2 had a small growth of only 5 W with heat load B input less than 4 W. It was just less than a 10% increase in threshold critical power for TGHP-1 and TGHP-2 (from 50 W to 55 W for TGHP-1; from 45 W to 50 W for TGHP-2). When heat load B increased from 4 W to 10 W, TGHP-1 increased rapidly (From 55 W to 80 W for TGHP-1), while TGHP-2 still increased slowly (from 50 W to 55 W for TGHP-2). It was almost a 50% increase in threshold critical power for TGHP-1 when heat load B increased from 4 W to 10 W. Moreover, TGHP-1 could be found to be better than TGHP-2 in the threshold critical power with heat load B increase, especially with high heat load B input. Therefore, TGHP-1 was chosen to carry on the following experiments. In conclusion, the addition of heat load B served as a thermal switch in both TSHP and TGHP, which could greatly increase the threshold critical power of a heat pipe system.
The distribution of steam pressure inside the heat pipes is demonstrated in
Figure 4. In this work, the addition of heat load B was introduced into the heat pipe systems, including TSHP-1 and TGHP-2. Therefore, it was quite different from the heat pipe system with only one single heat load, especially in the distribution of steam pressure inside the heat pipes. As shown in
Figure 4a, the steam pressure of P1 increased when the heating block was under normal working, and then the steam flowed along the heat pipe to the condensation section. Finally, the steam was condensed into water. The condensed water reflowed to the evaporation section through the capillary structure. It also meant that the steam pressure of P2 was low when the heat pipe was working normally. Moreover, the steam pressures of P1 and P2 were under dynamic equilibrium, but the condensed water was blocked in the condensation section when the heat pipe was under working overload status. It meant that the heat pipe could not be carried on heat dissipation in time due to large heat input, which led to the dry-out status of the evaporation section. As a result, the steam pressure of P1 was much higher than that of P2.
Figure 4b shows the distribution of steam pressure inside the heat pipe with two heat loads when the heat pipe was under normal and overload working status, especially for necking heat pipes such as TSHP-1. Before the heat pipe with two heat loads came into working overload status and the additional heat block was out of work, the distribution of steam pressure inside the heat pipe was the same as shown in
Figure 4a. When TSHP-1 was under overload working status without heat load B working, the condensed water was blocked in the condensation section. However, the steam pressure of P4 and P5 rapidly grew up after heat load B was on working due to the small volume of the necking section. Therefore, the blocked working liquid reflowed because the steam pressure of P4 and P5 was higher than that of P3. As a result, TSHP-1 came into normal working under higher heat load A input. Then, the steam pressures of P3, P4, and P5 returned to dynamic equilibrium. In summary, the addition of heat load B served as a thermal switch could improve the threshold critical power of a heat pipe system by increasing the reflow flux of the blocked working liquid inside the condensation section.
3.2. Dynamic Response Characteristics of the Heat Pipe Systems
Figure 5 demonstrated the surface temperature distribution of TSHP-1 after reaching the threshold critical power with different heat load B, including (a) 2 W and (b) 4 W. As shown in
Figure 5a, TSHP-1 was under normal working before 700 s for the tested points T1–T9 remained relatively stable. TSHP-1 then came into overload working status with the increased heat load A after 700 s. It could be seen that the surface temperature of T1 increased quite rapidly, which meant that the evaporation section was dried-out. However, the surface temperature of T1 recovered stability with heat load B input of 2 W after 1000 s. The stable status of TSHP with heat load B input of 2 W remained at approximately 250 s, and then the surface temperature of T1 increased rapidly to cause the dry-out status of TSHP-1. The stable status of TSHP recovered after heat load B input of 4 W, as shown in
Figure 5b. The surface temperature of T8 and T9 grew up when heat load B was input, while that of T1 decreased to the original state. Moreover, the surface temperature of the tested points was kept stable, which meant that the whole heat pipe system could maintain stability after heat load B input of 4 W. The surface temperature distribution of TGHP-1 after reaching the threshold critical power with different heat load B is shown in
Figure 6. The changing situation of TGHP-1, which recovered from the dry-out status to the stable status, was almost the same as that of TSHP-1. However, the minimum heat load B input of TGHP-1 was quite different from that of TSHP-1. When heat load B input was 4 W, TSHP-1 could be recovered to the stable status, while TGHP-1 would be recovered stable when heat load B input was at least 6 W.
From the above, it could be known that the heat pipe system would recover from the dry-out status to the stable status with a proper heat load B input. Furthermore, the recovery characteristics of TSHP-1 with heat load B input of 10 W after the surface temperature of T1 dried out to different temperatures, including 75 °C and 94 °C. As shown in
Figure 7a, the recovery time of TSHP-1 after being dried out was quite short (less than 150 s) with heat load B input when the surface temperature of T1 was 75 °C. It meant that the addition of heat load B played a significant role in adjusting the status switching of TSHP-1. Moreover, TSHP-1 recovered rapidly when the surface temperature of T1 was low. Therefore, TSHP-1 recovered to a stable status quite rapidly when the evaporation section was under the dry-out status for a short time. On the contrary, TSHP-1 recovered quite slowly because the surface temperature of T1 was high. It also meant that TSHP-1 was out of stable status for a long time. As a result, nearly 450 s was paid for the status recovery of TSHP-1 in
Figure 7b. The reason for the different recovery times of TSHP-1 with a different surface temperature of T1 may be that TSHP-1 under the dry-out status with different times caused different pressure distribution in the heat pipe, and it would greatly influence the rebuilding time of the stable status in the heat pipe because the imbalance pressure distribution of the heat pipe changed with time under the dry-out status. For example, it cost nearly 450 s for TSHP-1 to return to the stable status after the surface temperature of T1 was up to 94 °C. It could be seen that the pressure of the evaporation section was high enough to prevent the working liquid in the heat pipe from reflowing back. Additionally, the liquid slug might be formed in the condensation section. However, the liquid slug could be pushed back to the evaporation section with heat load B input of 10 W, even though it took a long time in this process. That was why nearly 450 s was the cost for TSHP-1 to recover to the stable status, which was approximately three times longer than that of status when the surface temperature of T1 was 75 °C. More detailed explanations about the pressure distribution of the heat pipe during different statuses can be found in
Section 3.1.
Figure 8 demonstrates the surface temperature distribution of TSHP-1 versus heat load A and heat load B. As shown in
Figure 8a, the surface temperature distribution of TSHP-1 with different heat load A from 45 W to 80 W under a certain heat load B input of 12 W was demonstrated in detail. It could be seen that the surface temperature along TSHP-1 from the evaporation section to the condensation section (from T1 to T6) showed a small difference before it dried out, except for the heat load B section (from T7 to T9). TSHP-1 would have good temperature uniformity performance with heat load A input less than 75 W under a certain heat load B input of 12 W. Because the heat pipe was under the stable status. However, the surface temperature of T7 to T9 was quite different from each other. The surface temperature increased successively from T7 to T9 due to the rapidly increasing steam pressure during the heat load B section. Moreover, the surface temperature of T1 and T9 would suddenly increase when TSHP-1 came to the threshold critical power under a certain heat load B of 12 W. It was worth noting that the surface temperature of the whole heat pipe slightly increased before TSHP-1 came to the threshold critical power. The surface temperature distribution of TSHP-1 with different heat load B from 0 to 12 W under a certain heat load A input of 45 W was shown in
Figure 8b. The surface temperature along TSHP from the evaporation section to the condensation section (from T1 to T6) showed a small difference for TSHP-1 under the stable status during the whole testing process, except for the heat load B section (from T7 to T9). The surface teperature still increased successively from T7 to T9 due to the rapidly increasing steam pressure during the heat load B section. Firstly, the surface temperature of T7 was higher than that of T9 without heat load B input, and then the surface temperature of T9 increased slowly versus heat load B. Finally, it exceeded that of T7. In summary, the main section of TSHP-1 (from T1 to T6) had good temperature uniformity performance when it was less than the threshold critical power. The addition of heat load B for TSHP-1 could cover a larger power range than that of a heat pipe with the same dimension. Moreover, the necking section of TSHP also played an important role in the steam pressure distribution adjustment. Therefore, different heat loads B could obtain the different threshold critical power of TSHP-1, and it also meant that low threshold critical power heat pipe could normally work under an overpower status with heat load B input. It also meant that TSHP-1 with heat load B was especially suitable for overclocking work situations.
The thermal resistance of TSHP-1 versus heat load A and heat load B was demonstrated in
Figure 8c,d. It could be seen that the total resistance of TSHP-1 kept on a low level before TSHP-1 was under the dry-out status when a certain heat load B input was 12 W. When heat load A was over 80 W, TSHP-1 was out of normal working for the total resistance grew up rapidly. Moreover, the evaporation resistance of TSHP-1 was a little higher than the total resistance of TSHP-1, and the condensation resistance of TSHP-1 was even lower than 0. The reasons for the above phenomena were that the steam pressure of TSHP-1 might be under a dynamic balance status. Moreover, the status could be speculated as follows: two ends of TSHP-1 were under high steam pressure, while the middle side of TSHP-1 was low with the increase in heat load B. Here, heat load B served as a thermal switch to adjust the steam pressure distribution of TSHP-1. Then, as a result, the surface temperature of the adiabatic section of TSHP-1 was slightly lower than that of the condensation section. Furthermore, the surface temperature of TSHP-1 might have been influenced by the thermal conduction of the shell and then caused the phenomena above. As shown in
Figure 8d, the thermal resistance of TSHP-1 with different heat load B from 0 to 12 W under a certain heat load A input of 45 W was demonstrated in detail. The overall trend was that the total resistance and condensation resistance of TSHP-1 decreased and then kept stable with the increase in heat load B, while the evaporation resistance increased first. The reason could be attributable to the addition of heat load B because the steam pressure of the condensation section was low firstly without heat load B input. Then, the condensation resistance was quite high compared with the others. As heat load B input increased, the balance of steam pressure was broken again. It caused the status where two ends of TSHP-1 were under high steam pressure, while the middle side of TSHP-1 was low. Therefore, the condensation resistance of TSHP-1 was even lower than 0.
The thermal resistance of TSHP-1 before and after it dried out is shown in the following
Figure 9a,b.
Figure 9a demonstrates the thermal resistance of TSHP-1 with different heat loads A from 10 W to 55 W, which meant that TSHP-1 was under a stable status. Additionally, heat load B was unworking. It could be seen that the total resistance of TSHP-1 appeared to have a U-shaped change with the increase in heat load A. When heat load A input was low, the evaporation section of TSHP-1 was not fully started up, which caused some working liquid accumulation. As a result, the evaporation section resistance of TSHP-1 was high. With the increase in heat load A, the working liquid accumulation gradually improved, and the evaporation section of TSHP-1 even dried out because the evaporation section resistance increased. Moreover, TSHP-1 obtained the minimum resistance with heat load A input of 35 W. The thermal resistance of TSHP-1 with different heat load B from 0 to 12 W under a certain heat load A input of 45 W is shown in
Figure 9b. TSHP-1 could reach a certain threshold critical power with a certain heat load B. Firstly, the total resistance of TSHP-1 increased with a low heat load B. During this process, the balanced steam pressure of TSHP-1 was broken, and it was hard for TSHP-1 to recover to a stable status with such a low heat load B. However, the total resistance of TSHP-1 was still at a low level, which was also in line with the normal operating range of heat pipes. The total resistance of TSHP-1 was then reduced with heat load B input from 2 W to 6 W. Finally, the total resistance of TSHP-1 increased with heat load B input from 6 W to 12 W. The reason for it might be that localized drying occurred during the evaporation section of TSHP-1 with a high threshold critical power of heat load A input. Therefore, the total resistance of TSHP-1 increased even with a high heat load B input because the steam pressure distribution of TSHP-1 might be under a delicate balance. At this point, the steam pressure of the evaporation section was a little higher than that of the condensation section. The thermal resistance of TGHP-1 versus different heat load A and heat load B was demonstrated in the following
Figure 9c,d. It could be seen that the general tendency of TGHP-1 was almost the same as that of TSHP-1. Therefore, the total resistance of TGHP-1 was a little higher than that of TGHP-1. Moreover, the condensation resistance of TGHP-1 was a little lower than 0. The reason for it might be that TGHP-1 still had a long section (about 30 mm) for heat load B with the same diameter as the main section along TGHP-1, and the liquid slug might have been formed easier without heat load B input. In summary, the addition of heat load B can adjust the steam pressure of TSHP-1 and TGHP-1 and thus increase the threshold critical power of TSHP-1 and TGHP-1. Moreover, TSHP-1 was more suitable for the real situation. Thus, it is very suitable for a real situation where the waste heat of a low-power chip can be utilized to drive a high-power chip working under the overload situation on the same circuit board, which can be considered waste heat recovery and utilization because a low heat load B input could obtain a high threshold critical power of TSHP-1.