AIR COOLING OF AN ELECTRONICS UNDER CONDUCTION, NATURAL CONVECTION AND FORCED CONVECTION

In recent years, as electronic gadgets have been more functional and become smaller, specialists have been confronted with challenging thermal issues. While power utilization has been required for faster processors, this always ends up with more generation of heat in electronic devices. Thermal and pressure behavior of air flow is significant to understand the cooling mechanisms while keeping the electronic components at desired performance levels. This project studies the thermal behavior of a powered object mimicking a small electronics component cooled at one of the following methods: conduction, natural and forced convection. A theoretical, computational and experimental study have been performed for all experiments. In the analytical study, models have been developed by means of Engineering Equation Solver (EES) software for all experiments and analytical work has been accomplished by determining thermal resistance networks, Grashof, Rayleigh, Nusselt, Prandtl numbers and heat transfer coefficients. In the numerical study, computational models utilized computational fluid dynamics tools to solve Naiver-Stokes equations. Results showed distribution of local temperatures, heat flow, velocity vectors and over-heated areas in different regions. For all simulations, mesh distribution has been optimized and boundary conditions were set on the same value with the experimental setup and conditions. Experimental study has been performed with both microscopic-infrared imaging technique and conventional micro thermocouples. CFD models were validated with the experimental findings for all three distinct cases. 0.6 Watts were applied in all experiments and temperature data were collected by a fully automated data acquisition system with the use of T type thermocouples and microscopic IR images. In addition, all measurements were recorded in steady state conditions defined by 0.1˚C variation in temperature values over 5 minutes. In experiments, test vehicle were inserted inside a wind tunnel with thermocouple and electrical wire connections with/without fan working inside the tunnel. Results were then compared in both CFD and experimental analyses and reasonable agreements were obtained. For instance, in conduction experiment, only 0.1% difference was recorded between simulation and experimental results; while in natural convection experiment, 1% and in forced convection experiment, 0.93% variation was observed.