View Single Post
Unread 01-18-2005, 11:00 AM   #12
BAcon
Cooling Neophyte
 
Join Date: Dec 2004
Location: UK
Posts: 22
Default

Quote:
Originally Posted by pHaestus
I do my best to eliminate all useless head loss in loops. It just makes no sense. You are thinking only about the head loss from 10" of tubing, but you also have to consider the head loss from those turns. Looking at them I'd guess the flow resistance to be somewhere between a 45 degree sweep and a 90 degree sweep. Take a look at this page and do the math:

http://www.procooling.com/articles/h...s_with_h2o.php

Call it K = 20f for every 90 degree turn or K=80f for your pretzel. Now consider that K = (L/D)f, so L/D = 80. Your tubing ID is 0.5", so your 10" pretzel has an effective length of 160" long. Now of course there would be SOME resistance from just the 10" of tubing (10/0.5 or 20f). So your little 10" pretzel has at least 8x the flow resistance it should if it were just a straight run of tubing.

Now does this actually make a difference in final CPU temperature? Sure it could. Is it going to be 4-5C? Nah probably not. I'm obsessive though and like to do things correctly
interesting perhaps a re plumb is in need then, i have a g4 on order so when that appears ill see what i can do with the tubing and see if the tline will fit at the top by the rad outlet

Cheers for the info dude
BAcon is offline   Reply With Quote