Watercooled custom case + passive?

[MonsterMoshi]

Hi, i'm in the process of building a custom case as part of a year 13 extended project. I was always planning on watercooling the build for aesthetic as well as functional purposes but now i'm thinking making it partly passive.
This is the original sketchup design:

And this is the progress up to now:

I designed all the CAD files using NX uni-graphics and all parts were very kindly laser cut bend welded and finished by KMF in Newcastle (nextdoor to Overclockers)
The rest of the work is up to me now...

So on to the point....
Two 240mm radiators will be positioned her as below. Air will flow straight through from one side to the other.

As i am quite pedantic about a lot of things, especially noise, i want the system a quiet as possible. The watercooling should ensure this when the system is in use for gaming but a lot of the time, the system is left almost idle downloading things or when streaming media. So i was planning on being able to make the system completely passive for some of the time. My plan to do this was to position a custom made passive radiator on the outside rear of the case where a plain cover would have gone.

The radiator would be made something like this:

This is certainly not to any kind of scale and i think the fin density here is wayy to high for passive.
The fluid would always flow through here first whether passive or not.

I need to think about what material to make it out of. I would just go straight for aluminium but i think i would likely have a problem with galvanic corrosion so copper is the alternative despite the cost. Brass may be an option however it is much less thermally conductive.
I will make it using a CNC mill so that is not an issue.

What do you think of this idea?
How effective would it be at dissipating the heat or is this likely to be a test it and see job?

 

[shaunwil]

Could always anodize the aluminum for corrosion resistance. Its a lot cheaper material to use. Remember a lot of the aftermarket car radiators are fabricated in aluminum. Try sourcing the same grade of aluminum.

Doesn't NX come with simulation for heat dispersion ?

Shaun

 

[shaunwil]

Looks extremely nice by the way

Shaun

 

[MonsterMoshi]

Didn't realise anodising helps resist corrosion, thats a good help as i was planning on anodizing it all black anyway. I will look in to it
Yes, NX has got the thermal and fluid dynamics although i haven't tried it out yet but i can't imagine it is the easiest thing to use but will certainly give it a good go when i'm back at school

Thanks

 

[eddyr]

An alternative option could be to mount a radiator on the top that would take advantage of convection to run passive instead of the 2 vertical rads. Under load 4 fans pushing air inward would provide pressure and airflow through the top rad for higher performance when needed, and should help minimise noise from the fans. This could be in conjuction with your designed passive radiator.
Are those fan intakes 140mm?

Running passive you do need lots of surface area and the water will carry a lot of heat (so your components will get hotter) before the temperature delta is large enough to allow equal dissipation from the loop.
As an example, under low load conditions and passive my 60mm 120.3 radiator, blocks, fittings get too hot to touch for more than a few seconds before my radiator is able to remove as much heat as the chips are adding to the loop. My radiator is top mounted so convection occurrs well.

 

[MonsterMoshi]

The intakes are for 120mm fans and the spacing is correct for a 240mm radiator.
I will be using a D5 pump so there will be high flow throughout the system which I believe is important for running passive and maintaining a consistant tempreture delta. Also I think, with a radiator designed to run passive instead of one intended for use with fans (so easy vertical convection and low resistance to air flow) then it will likely be more efficient.
I will also use a fan controller such as the Aquaero 5 so I can program fans to kick in when the CPU or GPU temp reaches above a certain level.

 

[eddyr]

I couldnt say about the radiator, a seach on XS forums might show up some informative threads from ppl who have built purely passive WCing, but id have thought the high surface area of fins and tubing of normal radiators works fairly well with convection but does not radiate well. I'd be very interested to know how it turns out, and the design you have looks to have lots of surface area.

Flow will help a little at the blocks but if you cant dissipate the heat your chips add to the loop, the loop temperature will climb to a degree where the delta between the loop and ambient air is sufficient that it can. Only at this stage will temps stop rising. At these higher water temps your chips will naturally run hotter regardless of flow. The blocks in particular and tubing get hot and radiate heat to your other components which is another factor to consider.

In terms of removing heat from the loop the effect of higher flow through the radiator will be minimal. The main thing is to prepare for is for things to get hot (but more than cool enough for the chips at idle/low load with power saving features on). This can take some time as the heat capacity of water is high.