Abstract

This study presents a simple method for predicting the temperature of the base-plate of a central processing unit (CPU) cooler. These coolers are used in desktop computers to keep the CPU at an optimum operating temperature. The technique discussed here combines a computational approach with an analytical solution. The cooler has a base-plate, four heat pipes, and many plate-fins that transfer the CPU heat to the surrounding air. The computational model selects a fin and adjacent airflow and obtains a correlation for the conjugate Nusselt number. The model’s analytical part depends on this correlation, and it predicts the base-plate and heat pipes’ temperatures. The parameters that affect the temperature are the flow Reynolds number, CPU power, and the heat pipes’ effective thermal conductivity. The ranges of parameters in this study are Reynolds number 100–1000, CPU power 50–400 W, and thermal conductivity ratio 1–100. While Reynolds number and thermal conductivity ratio decrease the cooler temperature, the CPU power increases the temperature. The heat pipes’ dimensionless temperature depends on Reynolds number and thermal conductivity ratio, but not on the CPU power. The CPU cooler’s thermal resistance is a function of Reynolds number and thermal conductivity and provides a convenient way to estimate the CPU temperature.

摘要

本研究提出了一种预测中央处理器 (CPU) 冷却器底板温度的简单方法。这些冷却器用于台式计算机，以将 CPU 保持在最佳工作温度。这里讨论的技术结合了计算方法和分析解决方案。冷却器有一个底板、四个热管和许多将 CPU 热量传递到周围空气的散热片。计算模型选择翅片和相邻气流，并获得共轭努塞尔数的相关性。该模型的分析部分依赖于这种相关性，它可以预测底板和热管的温度。影响温度的参数是流量雷诺数、CPU 功率和热管的有效导热率。本研究的参数范围为雷诺数 100-1000，CPU 功率 50-400 W，热导率 1-100。虽然雷诺数和热导率会降低冷却器温度，但 CPU 功率会提高温度。热管的无量纲温度取决于雷诺数和热导率，但与 CPU 功率无关。CPU 散热器的热阻是雷诺数和热导率的函数，它提供了一种方便的方法来估计 CPU 温度。但不是在CPU电源上。CPU 散热器的热阻是雷诺数和热导率的函数，它提供了一种方便的方法来估计 CPU 温度。但不是在CPU电源上。CPU 散热器的热阻是雷诺数和热导率的函数，它提供了一种方便的方法来估计 CPU 温度。