http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Kholi, Foster Kwame 부산대학교 대학원 2021 국내박사
The efficient heat management of thermal systems continues to present challenges for designers in different thermal sectors. As various traditional and conventional heat management approaches maximum performance, new improvements in existing and future trends are required. Phase-change heat management, using heat pipes, promises an alternative solution with efficient features: low cost, less weight, passive operations, and reliable designs. However, the factors governing heat pipes' functionality are multidisciplinary, slowing down a full understanding and use in real applications. The inherent thermo-hydrodynamics are highly sophisticated, complicating their performance prediction in the early design stage. Therefore, in this research, a renewed effort was made to fundamentally understand the factors that govern the wick-based and wickless heat pipes. The main goal is to understand and predict the thermal limits of heat pipes under different operating conditions. Based on the experimental results, an optimum heat pipe geometry (effective length and diameter), cooling conditions (cooling lengths, flow rates, and temperature), and different heat pipes wick types (conventional, homogeneous, and composite) structures could enhance the thermal limits and minimise the effects of different modes of inclinations (stepwise, regular, and irregular) in mobility applications. However, other design limitations, such as large weights of some wick structures and low thermal limits of the wick-based heat pipes, necessitated a further investigation into the wickless (pulsating) heat pipes. This study carefully assessed the governing factors regulating the wickless heat pipes from the existing literature and experimental database. A novel correlation was developed based on this assessment to predict the wickless heat pipes (pulsating heat pipes) performance over a more comprehensive operating range. The proposed correlation accounts for the effects of the heat pipes' geometrical features, dimensionless numbers governing the internal thermo-hydrodynamics, and operational factors. The new correlation found a polynomial and power-law relationship between the heat flux and the heat pipes’ inclination and fluid filling ratio, respectively. Previous correlations have limited thermal predictions within a narrow range of parameters. Compared to such correlations, the new semi-empirical correlation has an improved scope over a wide range of fluid filling ratios and PHP inclinations. The achieved accuracy of ± 30% and R2 of 0.8651 when predicting selected experimental data shows reasonably good agreement. The new correlation applies to different working fluids, geometrical aspect ratios, and heat loads. The new correlation with its flexible application range is expected to assist in faster and more enhanced thermal predictions as interest in PHPs grows. Despite the enhanced scope of the present correlation, the high uncertainty associated with prediction is considerable enough to be minimised. Regarding this goal, the Artificial Neural Network (ANN) was employed to minimise the prediction error using highly non-linear algorithms. This modeling approach reduced the error associated with the correlation with an R2 of 0.9965. By nature, the ANN models are black-boxes and challenging to interpret physically. However, with the ANN models' coupling with the semi-empirical correlation, it was possible to enhance the heat pipes performance prediction and obtain a physically reasonable engineering explanation of the underlying multiphase phenomena.