[SOLVED] Custom PC desk....Do I have to mount fans DIRECTLY to rad?

[Xsv]

I'm in the middle of designing a custom PC desk. Will be using 2-420mm rads. If rads are in sealed encasements (except for filtered intake / open exhaust ports), do the fans have to be mounted directly to rad? There would be no angles for air to flow around. Straight in / out encasement. I don't see how it would impede performance. It might even improve flow by removing center dead area of fan spindles that usually sit directly on rad. Thanks in advance.

 

[NightHawkRMX]

You need pressure through a radiator, so just having air blowing at it is not enough.

 

[Paperdoc]

There's another factor. For really good air flow you usually use several fans. Moreover, your best arrangement is to have nearly equal airflow IN through intake ports and OUT through exhaust ports. So, where do you mount those fans? Normally, at the ports. Then, at either intake or exhaust locations you also must mount your radiators. (2 X 420 MM? REALLY? MAYBE 2 X 240?)

Suppose, for example following your query, you were to mount intake fans on some ports, and then mount your radiators at the exhaust ports with NO fans on them. You certainly will get some air flow that way. But now imagine adding fans at the radiators to help push air from the case through them to the outside. You can grasp that this will increase the total air flow though the case (and the rads), improving cooling capacity. This latter situation would be called mounting the radiators in the exhaust position which uses somewhat warmed air inside the case to remove heat from the rads and send it outside. A similar scenario places the fans and radiator on the intake ports so that the air flowing over the rads is cooler (outside) air, and the heat they remove from the rads must travel though the case with the air until it reaches the exhaust fans at their ports. It is commonly held that using the latter arrangement (intake rads) is better for cooling the hot components serviced by the rads, and has little effect on cooling the other mobo components.

 

[Karadjgne]

Back In the early days of Watercooling, that same question was answered by shrouds. You attached the fan to the shroud, and the shroud to the rad. That allowed enough distance between fan/rad to about eliminate the dead spot, but not so much that static pressure was lost. It was usually 25-38mm tops. Martin's Labs did a lot of testing with such.

Anything further than @ 38mm and the loss of SP became noticable, as air will compress to the point where it exceeds the ability of the fan to prevent backflow.

https://www.amazon.com/Phobya-Shroud-Plexi-Acrylic-120mm/dp/B004CLFFIA

Yes paperdoc lol, a 420mm rad is 3x 140mm, Op plans on using 2x 420mm in the loop.

 

[Paperdoc]

Wow. If the plan us to use TWO huge rads, why try to limit the number of fans? MAYBE an option is to mount one rad at an intake, and one at an exhaust location, each with its full set of 3 fans. I guess that depends on whether that would provide enough case cooling for whatever the system is.

If the two rads are in SERIES in a single cooling loop, then there is an optimal way to position them for an intake / exhaust arrangement. If they are separate cooling loops, then it's less important.

 

[Xsv]

I really appreciate the responses, but I think my question was foggy. Let me try this way:

Two separate acrylic box type casings, each with:
1- 420mm rad with three 140mm fans each, complete with separate loop.
2- Rad & fans will be housed in a separate / sealed enclosure away from everything else.
3- Both enclosures will only have three 140mm intake holes on bottom end & three 140mm exhaust holes on top end.
4- Straight through air flow with no angles or other openings. Hoses will have sealed pass-through on side.
Equal intake / exhaust holes.

If the rad and fans are as 1 unit at top OR bottom of casing, there would be X airflow across rad. (A)

If however, I mount three 140mm fans at intake end of the acrylic box and mount rad at exhaust end of the acrylic box, in a sealed fashion, would X airflow and rad cooling results be the same as (A)? Just want to relocate fans to aid in noise cancelling.

Basically like a pipe with one way in, one way out. Straight shot. No bends, angles, or other equations.

 

[Karadjgne]

You are creating very large shrouds. In a nutshell. That would work, depending on the fans.

Think of it like this. Blow out hard, almost a whistle. Put your hand next to your mouth. Feel the static pressure going through your fingers. Hard right?
Now take a paper towel roll and do the same thing, hold that up to your face and blow through it. Exact same cfm comes out of your mouth, but the distance is so great, by the time it gets to your fingers, the pressure is already drastically lower.

It'll get to the point where you'd need a Lot more input, because the air in the shroud will need to be moved too, and air is heavy and compresses. If you had 3x Delta 8k rpm fans with massive SP, there might be enough pressure built up to force the air through the rads, but otherwise it'll be back flow on the fans, too much resistance. The rads might as well be passive.

 

[Xsv]

Thanks...exactly what i needed to know. Didn't think about the air in the actually box as well. Learn something every day. To answer Paperdoc, am using separate loops for CPU and GPU. The design was for performance, acoustics, and aesthetics. Already have the 2080 Ultra, working on Threadripper purchase. Want the RAM & lane count. Again, much appreciated everyone.

 

[Karadjgne]

For acoustics, have the fans interior pull, the blade sound will get muffled by the fins on the rad. For a push, I'd not use any type of shrouding except multiple 120mm seperates or you'll get bleed pressure going sideways as fans are not exact, you can/will have ± 5% ish between fans for a total of upto 10% difference in speeds and at 1800rpm, that adds up.

 

[Xsv]

For acoustics, have the fans interior pull, the blade sound will get muffled by the fins on the rad. For a push, I'd not use any type of shrouding except multiple 120mm seperates or you'll get bleed pressure going sideways as fans are not exact, you can/will have ± 5% ish between fans for a total of upto 10% difference in speeds and at 1800rpm, that adds up.

k...thanx for the added info