Heatercore's in parallel

[UberCool]

Basically I have a q6600 that is cooled by a Swiftech Apogee GTZ, MCP355 w/XSPC top, and I have 2 '77 bonneville heatercores (with 2 high flow S-Flex fans each). My question is do you think it is better to run them in series or parallel?

It appears to me that parallel would not actually benefit me in cooling because each HC would only be seeing 1/2 the flow. It would however lower the restriction of the loop. Series would benefit me in cooling assuming that the one HC can't get it down to ambient (I don't believe it can) but would add more restriction to the loop.

If no one has actually tested this before I may have to try both setups and see what I can find out (I may give both a try even if someone has). I do have a King flowmeter so I can see how flowrates are affected and I can judge temps off of the CPU reading.

[nikhsub1]

Basically I have a q6600 that is cooled by a Swiftech Apogee GTZ, MCP355 w/XSPC top, and I have 2 '77 bonneville heatercores (with 2 high flow S-Flex fans each). My question is do you think it is better to run them in series or parallel?

It appears to me that parallel would not actually benefit me in cooling because each HC would only be seeing 1/2 the flow. It would however lower the restriction of the loop. Series would benefit me in cooling assuming that the one HC can't get it down to ambient (I don't believe it can) but would add more restriction to the loop.

If no one has actually tested this before I may have to try both setups and see what I can find out (I may give both a try even if someone has). I do have a King flowmeter so I can see how flowrates are affected and I can judge temps off of the CPU reading.
You seem to know the details of running both ways, and you are correct. One other factor to think about that you did not mention was that with rads in series, one will always be less efficient than the other. The rad receiving the hotter water will be more efficient, the rad receiving the cooler water less so. With parallel rads you don't get this effect, both rads are equally efficient. FWIW, I ran a loop for a long time with 2 single fan HC's in parallel. With that said, Cathar mentioned once that he did do an informal test of 2 HC's in parallel and series, he found series to be a bit better. My suggestion is what you have mentioned, try both ways and see which is better in your setup.

[UberCool]

As I was thinking about it more today, I know that two rads in parallel should only cool about as much as one but I am wondering why. It seems to me that since with two in parallel the fluid will be spending twice as much time in each radiator you would see an increase in cooling. My best guess is that higher flow rates will cause more turbulence and transfer heat better between the water and copper but I have no scientific basis behind this. Does anyone know why this is?

[atomic ferret]

Sounds interesting, I'd like to see a comparison with temps. I'm considering buying a second '77 Bonneville heater core for my loop and would like to see the results.

[rally9x]

As I was thinking about it more today, I know that two rads in parallel should only cool about as much as one but I am wondering why. It seems to me that since with two in parallel the fluid will be spending twice as much time in each radiator you would see an increase in cooling. My best guess is that higher flow rates will cause more turbulence and transfer heat better between the water and copper but I have no scientific basis behind this. Does anyone know why this is?

I would guess that splitting and "unsplitting" the water will have its cost, esp. since most people will use plastic which I think is a little bit more restrictive.

I just got 3 HC's and 3 D4's so I was thinking about trying a crazy 1pump/1rad x 3 setup so I'll be interested to see the results here.

[UberCool]

Thanks for all of the input so far guys. I am going to be building a custom case this weekend to house my monstrosity so I should be starting my tests this coming week and we can see what I come up with. I will see what I can do about only testing it with one heatercore also so that we have a sort of "base" case to see how the other results stack up to it.

[Nightstar]

I did experiment a bit with parallel vs series flow through four 120s. I noticed slightly lower temperatures in a series/countercurrent arrangement YMMV. I think if your pump had a larger discharge nozzle than your rad's barb there might be an advantage to parallel.

I think the accepted theory is that decreased flow through the rads from parallel plumbing promotes a laminar flow reducing the efficiency of the rads.