Advantages of 2 rads?

[MasterOfPupets]

Well, I'm looking at taking my first step into Water-cooling, and this is what I'm thinking.

if I pick up a Thermaltake Tsunami case. This case has a front and rear 120mm fan. What I want to do is use 2 seperate 120mm Rads. One with air coming in from outside (the front intake) and one as my exhaust (the back fan/rad). Is there any disadvantage to this. Here's the loop I was considering.

Res - Intake Rad. - Pump - CPU - Exhaust Rad - GPU - NB - Res

I'd also have a second intake on my side panel, and a second exhaust on my PSU. Any advice on whether this would be good or bad?

 

[Ice Czar]

youd be heating the air in the case with the front rad, and lowering the effciency of the heat transfer of the rear radiator, since that air would be hotter, in addition the larger the loop and more convoluted the greater the restriction

http://www.procooling.com/articles/html/maximizing_flow_rates_with_h2o.php

an "interesting thread"
http://hardforum.com/showthread.php?t=770036&page=1&pp=20
but contains the info on differential temperatures I think you need to understand and some useful links

 

[MasterOfPupets]

Yes the air would be slightly hotter going through the second rad, it would still cool the water that was just used to cool the CPU. Basically, It'd cool the water more than not having one, and it would still provide an air exhaust for the case. Also, the flow restriction would be no less than running a dual 120mm Rad.

 

[Ice Czar]

heat transfer is all about differential temperature
and flow restriction in the rads isnt the only factor

 

[MasterOfPupets]

Well, I had a full reply written, but IE screwed up on me, and I'm not retypeing, so here's a short version.

I agree with you on the flow restriction, but I think it would be worthwhile for the convenience.

As for the temp, after reading all the links I could (some where broken) I came to the conclusion that there is not enough information to actually make a descision as to whether it would be worthwhile to re-cool the water. There's just too many variables. If you disagree, feel free to prove me wrong, but most of what you've posted is related information that, while all is correct as far as I can tell, doesn't prove anything one way or another.

 

[Ice Czar]

feel free to prove me wrong,.

wasnt trying to prove you wrong in the first place
any thermal solution depends on a great number of variables
I was just pointing a few of those out,
you reach a certain point where there is a diminishing return on investment
if your able to recognize certain pitfalls you might be able to avoid them with a design change

in this case the main variable would be the air temperature when it enters the case
if its cold enough then no problem, but if its not so cold, there will eventually be a temperature where it is unable to transfer all the heat for either the size of the interfaces or the airflow (you can increase either to address those)

basically your talking about drastically increasing the transfer interface

its a tradeoff between the ambient temperature of the air
the total interface area (in this case the rads)
and the amount of airflow
for the total heat that needs to be transfered

you can run a single rad 1/4 the size of what you have and with a really low airflow
if the computer is parked in the Antarctic

or you could employ the single rad and transfer the same heat as the proposed double
if you increase the airflow by some factor with a honkin big blower

or your proposed solution of doubling the heat transfer interface

I was simply pointing out that, because your probably altering the interior case temperature several degrees higher than the ambient temperature outside the case
that the "normal" heat transfer from components not water cooled in the case will be less
like the PSU, passively cooled elements like the Voltage Regulation Modules (which get real hot on my board) chips on cards
(I have a RISC processor on one of my RAID cards) ect. So everything used to be acceptable at a room temperature between lets say
up to 80F, with your setup thats going to drop some lets say 70F
of it might still be acceptable, even at higher temperatures

point is without some real sophisticated Computational Fluid Dynamics Modeling software
or trial and error you wont know what those parameters are. It might be acceptable then again it might not.

But lets say you had a dual rad but instead of with the first one at the intake, and the second as an exhaust,
they where both exhausted, all of a sudden the above becomes a moot point
it would be a good efficiency move

PS all the links checked out for me

.

 

[MasterOfPupets]

I see your point, and I like the idea of a dual exhaust. Problem then becomes moving enough air into the case for the airflow to stay high enough to be usefull. One way I could quickly think of achieveing this is by making my PSU intake air instead of exhaust it, but I would still probably need to mod in another blowhole (in addition to an 80mm on the window) to get the airflow high enough. A great idea none-the-less though. That's something I'm definately going to have to think about. It's nice to have someone who actually posts facts and lets people decide on their own, instead of forcing their oppinions on people.

 

[Ice Czar]

its all trade offs, consider this
if you had strategically placed intake ducts (they dont need to be powered the "draw" from the exhaust fans will generally take care of that) you can aim them at components or with airflow patterns that can assist "hot" spots, like certain chips, the pump body, lights ect

what would be ideal is to map the current temperature gradients in the box with a sensor (like a digidoc) to ID hotspots, and then work through various airflow patterns to address them, many members have gone as far as putting a plex or other clear temporary cover on a case and doing smoke tests to observe the airflow patterns

tuning either an aircooled or hybrid air-cooled\water-cooled system is never just a one shot deal (note the maximizing flowrates article as an example)

of course with either of these the real objective is a rig that will cool evenly to the ambient temperature of the room, as you cant go lower
and with the least investment in power (pump fans ect)

in an extreme setup that alters to the lowest temperature possible or a set temperature with the lowest investment in power

other criteria is maintence, noise level ect

Good Luck

 

[cloaked]

not to be one of those that 'forces their opinions on others' but i will have to agree with ice czar here. [prepare for gross but somewhat effective oversimplification here]. not to mention the (possible) loss of pressure by adding the extra rad and tubing to/ from it, the only thing in a watercooling system that is actually getting rid of the heat is the rad. this is also all passive, so the temperatures of the rad (aka the water in your loop) will in theory try to match the temperature of the passing air. although the difference between outside and inside air will not be _too_ great (depending on amount of hdds and other aircooled components), the back radiator will be recieving hotter air and although will most likely not be gaining heat, will probably not be dissapating enough to warrent it being there. a better solution would be putting the second rad by an intake or just getting a single bigger one if this is possible inside your case.

 

[rogue_jedi]

you could just duct both rads to the outside of the case (exhaust on the front one to the outside, intake on the back one to the outside) of course, it sounds like doing that would effectively remove the intake/exhaust from your system. you could probably get decent improvement over no ducts without losing all airflow through the case by just ducting the back one. also, you should play with the order of the loop to get the water at the maximum temperature differential at both radiators, to increase performance, i.e. put the hotter water from the cpu through the rear radiator, and then from there through the pump and then the front radiator (or put the pump in between the cpu and the rear rad, whatever works best) doing this properly will require some toying with the layout to make it work properly. other than that, the best advice i can give is to try it out and see what works best.