近年來，電子通訊設備處理速度與傳輸速度大幅提升，隨之而來的高溫已經成為設計人員必須面對的課題。本文研究之通訊設備是在自然對流情況下，僅靠流體本身之溫度差進行熱傳遞，過去憑著經驗與試誤方式進行設計，往往浪費大量時間與成本，因此需要有系統性的理論與實驗進行比對，並在有限時間內找出散熱模組之效能最佳化策略。 本文研究通訊設備系統內之有限空間散熱效能，利用熱流分析軟體FloTHERM 12並配合田口方法進行模擬規劃，品質特性為溫度低(望小)並選用直交表L9(34) 四因子三水準準位進行模擬，依照過往產品經驗選定以下四因子：散熱器表面塗層之熱輻射率(A)、導熱矽膠片之導熱係數(B)、散熱鰭片數量(C)、散熱鰭片高度(D)。依照田口法實驗設計進行模擬並得出9組數據，透過Minitab 20統計軟體以品質特性望小為目標進行統計，得出本實驗散熱最佳因子水準組合為：A3B3C2D3，也就是散熱器表面塗層為石墨稀奈米碳、導熱矽膠片之導熱係數為5 W/(m·K)、鰭片數量為10 pcs、鰭片高度為45 mm。研究結果發現關鍵因子為導熱矽膠片，其S/N比為0.34，各水準溫度差異達到3.63°C，然而在主要電子零件其S/N比提升至0.56，各水準溫度差異達到5.85°C；在進行熱模擬分析與實驗量測數據比較，結果誤差為3.01%，因此熱模擬分析有相當程度的可信度，在產品研究與開發過程中若搭配田口方法與Minitab統計分析，將可快速且有效的取得最佳設計方案。 ;The heat bring by the process and broadband speed of electronically device raised substantially recent years, is becoming the task every designer’s nightmare. Designed upon the past experience, improved along lots of trial and error to achieve the final target normally come with extensive time and cost; The scope of this research is to analyze and discover the best strategy for maximize the performance of the thermal solution within the limit of time with systematic trial and theory. The simulation is setup and planed base on Taguchi Methods and using the heat flow analyze tool, FloTHERM12; The quality characteristics is low temperature (smaller-the-better) and the orthogonal array table L9(34), target the factors: heat radiation (A), thermal conductivity of thermal pad(B), quantities of the Fins(C) and height of the Fins(D). The optimal parameters of the thermal solution is: A3B3C2D3, heat sink coating with graphene, thermal pad with 5 W/(m·K) conductivity, contain 10 fins with height 45mm. The research results found that the critical factor is the thermal conductive silicone sheet, its S/N ratio is 0.34, and the temperature difference between each level reaches 3.63°C. However, in the main electronic parts, the S/N ratio is increase to 0.56, and the temperature difference between each level reaches 5.85°C. The relative error between simulation result and experiment is 3.01%; this proved the thermal simulation is creditable. The combination of the Taguchi Method and Minitab statistics analyze applied during the product development, provide the design of thermal solution better accurate and consuming less time.
