Designing a High-Heat Flux Test Rig for Intel Processors

[Cyrus]

Every semester, at least one of my classes involves a project. For Mechanics of Materials, we made a beam, performed an analysis and tested in a universal bending machine. For Thermodynamics, we had to design a HVAC system that provided Heating/Air Conditioning, dehumidification, a cool water supply, and power using microturbines. For Fluids, we had to make a working model of a Urinary System and perform an analysis on the flowrates and use the non-dimensional correlations to scale the findings back down to a human size bladder. This semester, we had to make a test rig for an Intel Chip. Here is the project statement:

ENME 332: Heat Transfer Design Project Fall 2006

Future electronic systems and power electronics will require increasing use of high-heat flux removal technologies-see the attached NY Times article on the thermal problems facing electronics. High heat flux thermal designs are necessary to maintain lower operating temperatures, which increases the reliability of components and can result in higher performance. Use of liquid cooling will become unavoidable as the power dissipation levels increases in future electronic systems. Possible liquid cooling technologies include singlephase liquid cooling in microchannels, immersion flow boiling, spray cooling, jet impingement cooling, thermosyphons, and heat pipes. We wish to build a test rig to measure the heat removal capability of various cooling technologies such as sprays and jets. It is your job to design a heater that can remove heat fluxes up to 2000 W/cm2 from a 1 cm x 1 cm area. The fluid striking the surface is at a temperature of 20°C, and the heat transfer coefficients can be up to 400,000 W/m2-K. One design that has been used in the past to test much lower heat fluxes is shown on Figure 1. One or more cartridge heaters are placed into a block of copper, and the heat flux is determined from the temperature gradient in the region below the cooled surface measured using thermocouples spaced a known distance apart. Some aspects you should consider in your design are 1). temperature limits, 2).accuracy of your heat flux measurements and instrumentation, 3). Cost. Your design should use readily available parts (i.e., no materials like “unobtainium”). Please submit a report describing your design (5 pages max., single spaced, 1” margins, Times New Roman) including designs considered, dimensions, heat transfer performance, measurement accuracy, a list of materials, and cost. You may work in groups of up to three—you should choose your groupmates from your section. The project is due Wednesday, Nov. 29th in lecture.

The hardest part about this project was that it was completely open. Anything goes, so long as you can provide up to 2000W of power. As my professor said, it's deceptively simple at first sight.

 

[Cyrus]

Here is the NY Times Article:
http://www.yousendit.com/transfer.php?action=download&ufid=1426A8E4107BDEF6

Here is my Report:
http://www.yousendit.com/transfer.php?action=download&ufid=F036A0E1224D977B

Thanks to Astronc, Gokul, and Marcul for helping me out.

 

[dichotomy]

I'm actually quite interested in this and the links don't work...

"File not found

The link you have clicked is not available.

Use only the link that was generated by YouSendIt and emailed to the recipient(s)."
:(