三維複合微米柱陣列表面之池沸騰臨界熱通量

Abstract

對於電子設備能量之需求將隨著工業技術的進步而增加，因此熱管理的可靠性與防止大功率之電子設備的損壞是非常重要的。然而在許多散熱方式中，池沸騰熱傳是最有效的方式之一，因為在相變過程中會消除大量潛熱。 本研究將運用三維複合微米柱陣列結構進行池沸騰熱傳實驗，而三維複合微米柱陣列結構為其表面具有上下層高度差，上層為疏水層控制成核位置，下層為親水層高毛細力之微米陣列結構。因此本研究預計三維複合微米柱陣列表面將達到液流與氣流的分離，並消除流體不穩定性之限制。本研究將使用高速攝影機觀測表面沸騰現象及紅外線攝影機運用於觀測表面之溫度分佈並進行成核密度量化與蒸氣覆蓋面積之分析。再者，本研究將改變測試端樣本之加熱方式，以此來消除邊界效應及獲得其維複合微米柱陣列結構真實臨界熱通量。 由實驗結果發現其臨界熱通量值與三維複合表面之高度差無關，其原因推測為其高度差差距太小。在文獻中，基於流體不穩定性之臨界熱通量模型之蒸氣覆蓋面積比率常利用貼合因子來預測，因此本研究將利用紅外線攝影機分析出三維複合表面之蒸氣覆蓋面積比率來預測理論模型。在蒸氣覆蓋面積比率結果發現其三維複合表面所計算之值極為相近，其平均值為0.215  0.003。三維複合表面之基於流體不穩定性理論所提出之理論臨界熱通量模型值約為220 W/cm2，而三維複合表面獲得平均臨界熱通量值為220.36  7.59 W/cm2，兩者結果接近吻合。再者，二氧化矽平表面之實驗值為110.62  1.6 W/cm2 與基於流體不穩定性理論所預測之理論臨界熱通量值(115 W/cm2)也相近，而二氧化矽平表面蒸氣覆蓋面積比率值為0.11  0.039。由以上結果說明兩者表面皆受到流體不穩定性所限制。不過三維複合表面與二氧化矽平表面相比，有較高之臨界熱通量，三維複合表面最高之臨界熱通量為237.85 W/cm2相較二氧化矽平表面有111之提升。其推測原因為(1)三維複合表面有較小之平衡接觸角，表面有較高之潤濕性，(2)親水端之小柱結構所提供之毛細力將有效潤濕表面。這表面三維複合表面可在較大尺寸之蒸氣覆蓋面積下才會產生乾點進而燒毀，因此三維複合表面在臨界熱通量下獲得較高之蒸氣面積，並提升臨界熱通量值。 而熱傳係數中，三維複合表面最高之熱傳係數為12.69 W/cm2-K相較於二氧化矽平表面有355之提升，其實驗推測原因為三維複合表面之成核密度較二氧化矽平表面高，使得熱傳係數有較高數值。然而改變三維複合表面之疏水層與親水層面積比率並無使其成核密度有無明顯變化，推測原因為設計三維複合表面之親疏水面積比率差距過小。綜合以上結果，三維複合表面與二氧化矽平表面相比，臨界熱通量與熱傳係數皆有所提升。

The power density of electronics increases with the need for industry advancement. Therefore, thermal management is important to improve reliability and prevent the permanent failure of high-power electronics. In many heat dissipation methods, pool boiling is one of the most effective due to the removal of a large amount of latent heat from the phase-change process. This work conducted pool boiling experiments on the three-dimensional (3D) hybrid micropillar arrays. The 3D hybrid micropillar array has a height difference between the hydrophobic top and hydrophilic bottom pillars. The hydrophobic top pillars are designed to prompt nucleation, and the hydrophilic bottom pillars can supply a large capillary force. Therefore, the 3D hybrid micropillar array can potentially achieve the separation of liquid and vapor flows, eliminating the mechanism causing hydrodynamic instability. A high-speed camera was used to observe bubbling dynamics, and an infrared (IR) camera was used to measure the surface temperature distribution and quantify nucleation site density and vapor coverage ratio on the surfaces. Furthermore, the pool boiling test section is revised to reduce the edge effect and to obtain the actual critical heat flux (CHF) values of the samples. In the literature, the vapor coverage area ratio at the CHF point in the hydrodynamic instability model is often adopted as a fitting factor. In this work, the vapor coverage area ratios at the CHFs on the 3D hybrid surfaces were measured using the IR camera. Furthermore, the effect of height difference between the hydrophobic and hydrophilic regions was investigated. The CHF values on the 3D hybrid surfaces were independent of the height difference. This is presumably because of the small height difference assigned on the 3D hybrid surfaces. The vapor coverage ratios at the CHFs on the 3D hybrid surfaces with various height differences were also similar, and their average value was 0.215  0.003. The theoretical prediction of the CHF using the hydrodynamic instability model with the measured vapor coverage ratio was 220 W/cm2. The experimentally obtained average CHF value on the 3D hybrid surfaces was 220.36  7.59 W/cm2. The experimentally obtained CHF on the 3D hybrid surface agreed well with the theoretical prediction. Moreover, the obtained CHF on the plan SiO2 surface was 110.62  1.6 W/cm2, which also agreed with the CHF prediction (115 W/cm2) using the hydrodynamic instability model with the measured vapor coverage ratio of 0.11  0.039 on the plain SiO2 surface. These results suggest that the CHFs on the 3D hybrid surfaces and plan SiO2 were both limited by the hydrodynamic instability model. The 3D hybrid surfaces had a CHF higher than that on the plan SiO2 surface. The highest CHF on the 3D hybrid surface was 237.85 W/cm2, which was approximately 111% higher than that on the plan SiO2 surface. The larger CHF on the 3D hybrid surface is presumably because (1) the 3D hybrid surface had a higher wettability, and (2) the hydrophilic pillars also provided a large capillary force to rewet the surface. Consequently, the 3D hybrid surfaces would not generate dry points until a large size of vapor layer covered the surfaces. Thus, a larger vapor coverage area was obtained at the CHF on the 3D hybrid surface, and therefore, a higher CHF value was obtained on the 3D hybrid surface. Moreover, the highest obtained heat transfer coefficient (HTC) on the 3D hybrid surfaces was 12.69 W/cm2-K, which was 355% higher than that on the plan SiO2 surface. The reason for this is that the nucleation site density of the 3D hybrid surface was higher than that of the plan SiO2 surface. However, there is no significant difference in the nucleation site density when changing the hydrophobic/hydrophilic area ratio on the 3D hybrid surface. It is possibly due to the minor variation in the hydrophobic/hydrophilic area ratio among the 3D hybrid surfaces. Nevertheless, the 3D hybrid surfaces can effectively improve the CHF and HTC compared to the plan SiO2 surface.