The Optimized Heat Pipe Arrangement on a Flat Plate with Taguchi Method and Genetic Algorithm

• Main Authors: MIN-FANG SUNG, 宋旻芳
• Other Authors: TING-HSIEN LAN
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/57808568786721551405

碩士 === 中華技術學院 === 飛機系統工程研究所 === 97 === The current electronic products on market are reaching a high capacity and high performance, and hence which generates a large amount of heat and must be dissipated within a finite space. The heat dissipation problem attracts a lot of attention from both the researchers and the industry. The heat pipe (HP) is one of their choices. In this study, the simulation software CFD-ACE+ was used in studying various heat generation 15W、45W、75W in two different cases, which consists of a single HP case and a dual HPs case. In the single HP case, both the constant width and the constant area of the HP are analyzed. But in the dual HPs case, only the constant width case was analyzed. The inclined angle and its width of the HP are the chosen parameters in this study. The selected inclination angle and the widths are 18°, 30°, 45°, 57°, 60°, 67° and 70°, and 3mm, 4mm, 5mm and 6mm, respectively. The maximum temperatures Tmax of the heat source base plate were calculated. In the single HP case, the calculated data set is fitted into a polynomial by using the least square curve fitting technique. Differentiating this polynomial and set it equals to zero, the minimum Tmax can be found. This predicted value was then compared with the confirmed calculation result. It shows that the best heat dissipation effect occurs at the width of 6mm, and the HP inclination angles are all about 62.6° for the various heat generation cases. When the HP is of a constant area of 240mm2, their optimized inclination angle are about 65.9°. In the dual HPs case, there are four parameters; two widths and two inclination angles, are involved in the analysis. The Taguchi method with the L9(34) orthogonal array was used to predicted the optimum design. A different method, by integrating the Neural Network and the Genetic Algorithm, was also used to perform the optimum design. It shows that the integrated method can have a better design than the Taguchi method. The optimum design is in the case of both the two HP widths are 6mm, and their inclination angles are in 37.5°and 72.5°, respectively.