Heat Transfer, Год журнала: 2024, Номер unknown
Опубликована: Дек. 2, 2024
ABSTRACT Effective heat distribution in electronic circuitry is essential to improve the performance and life of components such as chips. This study presents a numerical analysis transfer on substrate board populated with an array discrete sources, assumed be placed horizontal air channel for forced convection cooling. The packages performed, taking into consideration effect thermal contact conductance (TCC) between source (chip) board. dependence temperature Reynolds number at inlet heating power from investigated velocities ranging 0.6 1.4 m/s observed significant. Temperature coefficient are systematically increase dissipation source. Two configurations—inline staggered—are analyzed, staggered configuration showing superior cooling performance. improvement attributed fact that arrangements expose fewer sources pre‐heated before it exits system. Additionally, location reaching highest found highly dependent TCC bonding material substrate. A hybrid optimization strategy employed, by combining Artificial Neural Network (ANN) Genetic Algorithm (GA) optimizing sources. ANN used predicting distribution, subsequently followed GA minimize maximum attained generating varying other control variables like thickness, velocity, generation. thickness layer varied 0.225 0.271 mm generation 1000 2000 W/m 2 . Among them, important parameter controlling optimum results obtained proposed compared simulation reasonably close.
Язык: Английский