A friend of mine is writing a PhD thesis and needs info about semiconductor cooling. Where is a good place to look, especially for information about how humidity might affect some cooling equipment?

Ask a process engineer :D

Humidity does have an effect on cooling. The higher the humidity, the more energy that the cooling air can absorb before it gets hotter. However, in the overall scheme of things, the effects are minor.

The major thing that contribute to good cooling is a strong temperature differential (ie using the coldest air you can find), getting a high air velocity over the item being cooled, and also a high mass-flow of air across the item to be cooled.

I suggest that you just do a search on google (for example) for heat transfer.

Note that I believe that for "serious" semiconductor cooling, use is also made of oil etc, instead of just using air.

ah, liquid nitrogen is the best solution for anything, unless you can get it to freeze....

I'd suggest that if your friend wants to really earn his PhD, He look for the info on his own. :)

Indeed. One of the first things I learned in College was that the internet is one of the worst sources for legitmate information on any subject you wish to write a college grade paper for. Get back in that library and find articles from Science Journals, hard research data, and books about semi-conductor cooling. Believe me, if I've found a 1500 page book about the contents of Portland Cement, there has to be a bunch of books about cooling computer systems.

Cooling of computer systems may not really be the best thing to research - I still say to actually research the principles of Heat Transfer.

Heat Transfer was one of my worst classes, and my least favorite in my junior year. There are numerous engineer handbooks that have distilled info on heat transfer (in part), plus the classical treatements at various levels. In graduate school, I had to do a lot of research on various things, and that was long before the era of the Internet and easy accesss to east coast reference material (I was in the south). Any university these days should have a decnet library I would think. Both my undergraduate and graduate engineering schools maintained their own technical reference library. These are how I'd research material needed for a Doctoral Dissertation.

BTW, A thesis is for a Master's Degree, and the difference is that a Thesis need not be original thought; a Dissertation must contribute new thought significant to the field of endeavor. A Thesis can simply rehash the work of others; Disertations cannot (well, they are not supposed to, but some schools are sloppy or of low reputation).

The tinkerers who run their computers at higher than rated speeds, do it by cooling. Overclockers (http://www.overclockers.com/) has some very active discussions, applied and theoretical, and info on every means of computer cooling ever tried.