# from Russell Johnson on Thursday 03 January 2013:
>> Mineral oil has a specific heat of ~.4, water, 1.0. So you would
>> have to have 2.5 times as much mineral oil to accept the same
>> amount of heat and with the same rise of temperature.
>
>The difference I've seen in mineral oil cooling is that the whole PC
>is submerged, not just a small radiator on the heat generating
>components filled with water. At that point, you have many times as
>much oil as you would ever have water.
Unless the oil is circulating, specific heat won't matter once the
system comes to a steady state. With the system immersed, the driving
factors would be the heat conductivity and surface areas of the
components and heat sink. Whereas, with circulating water, you are
concerned with specific heat and flow rate. It all just reduces to heat
transfer, but how you calculate it is apples to oranges (and persimmons
in the case of the still air insulating around every un-piped thing in
an exclusively water-cooled case.) I'm thinking we should skip this
debate and move on to electrically non-conducting (and maybe non-
flammable) refrigerants. Just spray the liquid across the components
and allow the phase-change to take the heat away. :-)
--Eric
--
---------------------------------------------------
http://scratchcomputing.com
---------------------------------------------------
_______________________________________________
PLUG mailing list
[email protected]
http://lists.pdxlinux.org/mailman/listinfo/plug
