Heat capacity and thermal conductivity are different as well. I would bet that the thermal conductivity in water is slow compared to CU. Otherwise you could turn your pump off and the heat radiator would still work like a giant heat pipe.
Heat capacity I would assume defeats the purpose of a thermal conductor, in fact this give me an idea. It would be better to find a fluid with a much lower heat capcity than water, b/c all the high heat capacity means is that it takes a long time to heat it, but once it is hot it no longer makes a difference (the steady state temperature is determined not by capacity, but by conductance).
Anyway
Heat capacity in Celsius/J/g
Aluminum 0.902
Copper 0.385
Water 4.179
Air 1.01
So anyway It seems that if we found a liquid with a lower heat capacity but higher conductivity it might actually work better.
http://hypertextbook.com/physics/thermal/conduction/
(~300 K except where otherwise indicated)
k (W/m·K)
aluminum 237
copper 401
mercury 8.34
gold 317
carbon, graphite (∥)
1950
carbon, graphite (⊥) 5.7
carbon, diamond 895
silver 429
water, liquid(373 K) 0.679
So we need graphite heat sinks really
And btw that means if it is in the parallel direction, for those that do not know graphite is basically a series of hexagonal sheets, and this leaves out an electron for each hexagon that bonds to the sheet next to it, thus in the parralel direction (along the sheet) all kinds of great properties are acheived, but in the perpiducular direction things are not so rosy.
And yes mercury looks like it would make a great liquid till you died a horrible death
Maybe if nothing else pure ethanol could be used, I wonder personally whether I could make a heat pipe dype system on a large scale I actually kind of want to try it. Ethanol or something would boil in the heat sink and the vapor would condenst in the radiator. If you had a vertical radiator I bet it would work (liquid collects at the bottom and runs into CPU water block to boil again). Sounds fun till it explodes I guess.
