A little help with parallel rads...

[nibble]

Well after seeing a little discussion here recently on running parallel radiators, I'm convinced to give it a shot. I've been getting not too spectacular temps with my dual MCR220's, D-tek fuzion, MCW60 an DDC2 it was recommended by serialk11er that I do parallel originally but I wouldn't listen. My question is do the lines to each radiator need to be exactly the same, and if so could a shortage in length of the input to one rad by countered by an increase in length of the output of the same rad? Any help, tips, etc would be appreciated.

[NaeKuh]

Well after seeing a little discussion here recently on running parallel radiators, I'm convinced to give it a shot. I've been getting not too spectacular temps with my dual MCR220's, D-tek fuzion, MCW60 an DDC2 it was recommended by serialk11er that I do parallel originally but I wouldn't listen. My question is do the lines to each radiator need to be exactly the same, and if so could a shortage in length of the input to one rad by countered by an increase in length of the output of the same rad? Any help, tips, etc would be appreciated.

well i can be wrong here so dont take me 100%

But from my limited understanding about fluid dynamics. You need to make sure the lines going from the pump -> each radiators -> block. is in essense the same.

So that means your intakes to both radiators need to be the same length, and the output on both radiators need to be the same.

My easiest recomendation is try to get a Y as close as possible to both the intakes and output on that radiator so you dont need to worry that greatly on = tube length.


Remember how fluid works, if one side has more backpressure, water will just go the other direction. Then you will get worse temps then running them a series.


PLEASE try to take accurate messurements when you do this. We are all curious on how it will turn out.

Some good Tests would be a 20min TAT burn on stock and OC settings. Both series and paralell.

Also try to get your ambients as close to each other in both tests. If not please record the ambient temps. You can use a regular thermoter. The ones you put in your mouth when your stick, and stick it up front of your radiator if you dont have a digital one.

We just need the delta of ambient to idle temps, and load temps to calculate it mathmatically on if its actually benifical.

Of course i wont be doing it, but i know someone that will. :T


I say those are probably the biggest variables, since everything else will be essentially the same, and we can just cancel them out. You should collect coolant temps of both in and out rad, however, i wouldnt ask that much out of you unless your XS name was Nikhsub1, or Petra. :rofl:

[serialk11r]

I believe that the tubes/bends in the tubes to each rad SHOULD be as close to identical as possible optimally, but really, a little extra tubing, a curve here or there, doesn't really matter, just keep them approximately the same. The pressure drop of 1/2" tubing is really small anyways.

[nibble]

What about this
http://ocidb.com/albums/userpics/10058/parallelrads.JPG
I'm thinking that total resistance would be the same as tubing length/bends would be equal in the inlet and outlet although in different places, I'm probably wrong though. If it wouldn't work I could work around it, this would make life easier is all.

[serialk11r]

Yea that would work. It really doesn't matter THAT much because even like 6 inches of tubes doesn't add a whole lot of restriction.

[nibble]

You should collect coolant temps of both in and out rad, however, i wouldnt ask that much out of you unless your XS name was Nikhsub1, or Petra. :rofl:
Nope I'm no Petra or Nikhsub1... yet :D. I'll gather as much data as possible but I don't have a lot of tools at my disposal, so probably just temps of everything.
Yea that would work. It really doesn't matter THAT much because even like 6 inches of tubes doesn't add a whole lot of restriction.
Cool I'll go ahead ASAP so, just got to get an order in to Petra, hopefully he'll have stock of what I want.

[sc00p]

I just had to "recommend" (<-haven't anything to compare to) serial rad setup... but the way like... gpu->1. rad->cpu->2. rad->...
...I do it with MCP600 and it's fine (I do not dare to say, great ;)... I should do scientific comparison/data logging before that LOL)

[nibble]

I just had to "recommend" (<-haven't anything to compare to) serial rad setup... but the way like... gpu->1. rad->cpu->2. rad->...
...I do it with MCP600 and it's fine (I do not dare to say, great ;)... I should do scientific comparison/data logging before that LOL)

Well I've tried series with not great results so I'm gonna give parallel a shot.

[sc00p]

Well I've tried series with not great results so I'm gonna give parallel a shot.
Well that's good you're going to do that with identical rads :cool:
My experiment in the past wasn't any good cause of different rads :shakes:
BTW: did you have your loop like the way I described when rad@series ?

[nibble]

BTW: did you have your loop like the way I described when rad@series ?

No I had pump->rad1->rad2->cpu->gpu->res->, that may have been part of the problem.