Computer chassis cooling is a problem most well associated with high performance computing, which is commonly applied in simulation computing, servers, 3D media rendering, and in gaming PC’s.

Today’s cutting edge computer hardware can cumulatively consume over a kilowatt of power, all of which is likely converted into heat within the chassis.  Here are some average thermal envelope specifications for some middle-of-the-road computer hardware you might find in performance geared desktop computers:

  • Central Processing Unit: Phenom II X4 955 Black Edition – 125 Watts, ($140 USD)
  • Graphic Processing Unit/Card: GeForce GTX 460 – 160 Watts (~$190 USD)
  • Memory: ~11 Watts Per Stick (non-specific, DDR3 RAM)
  • Hard Drive: ~11 Watts Per Drive (non-specific SATA HDD)

The possible combinations of hardware you could include in a desktop computer build are massive and the specifications are always changing, but that is not entirely critical to this study.  I have simply used the above listed power consumption / thermal envelopes as heat sources.

Fan placement is typically designed such that airflow is directed in through the front, sides, and top of the case – over the critical components such as the motherboard, the graphics card(s), and the disk drives – then out through the back of the case.  Other methods include applying constant inward airflow from all fans to generate pressure within the case that forces air out through perforated ventilation panels around the case walls.  The primary objective of this project is to determine the optimum method of air-cooling the case, and what types of fan layouts will optimize the case temperature and overall cooling performance.  Factors in this design optimization include:

  • Number of fans used?
  • What size fan would work best?
  • What is the optimum placement location for the fans?
  • Front, side, top, bottom, or back?
  • Should the fan blow air in or out?

To accomplish this goal, I have studied the effects of various air cooling methods by varying fan configurations and orientations and performing a computational fluid dynamics and coupled thermal analysis on a model of the design alternatives.  This was done using the NX6 Advanced Simulation Package.  You can download the full report here with the analysis and results here.

This project was my first self-motivated design project with computational analysis.  It was a requirement for my MAE 477/577 CAE class, but the project was allowed to be whatever we wanted it to be.  I chose computer cooling because I’ve been interested in using commercial CFD software ever since taking a course in the fundamentals of Computational Fluid Dynamics and Numerical Methods.  The results of this project are not refined, and are not perfectly accurate, and the analysis could use improvement.  However, the learning experience from this project was extraordinary and I plan on continuing to develop this newly acquired skill in the future.