Counter-flow dual loop vs standard dual loop.

I'm inevitably going to add a second loop to my set up (I have a DIY chassis with plenty of space for pretty much anything) and I'm wondering if I should add a standard second loop, or I had another idea. What I'd really like to do is use heat exchangers (radiators designed to exchange heat from one loop to the other to keep dual loop systems at similar temps) and then use a VERY high rate flow system with the new loop with the liquid flowing in different directions on the two loops.

Physics and logic dictates this should have really good results, but I'm just here wondering if anyone has ANYTHING like this, and if so what are your results/experiences.

Depends exactly what your goals are for how you will want to go about what you have in mind... But the easiest solution for keeping dual loops at the same temp is using a shared reservoir. Then if you have radiators in both loops, both radiators will be cooling the entire system.

Sounds like a engineers explanation for a series loop.

All needs done is to get the manufacturing to make the parts.

In a CPU/GPU loop, when the CPU is working hard there is extra rad due to the raddage for the GPU also. So the CPU uses the GPU extra raddage to stay cooler than if just a CPU loop.

The same if the GPU is stressed. There is extra raddage due to the CPU being in the loop, so the GPU stays cooler.

That is a naormal CPU/GPU loop which is more than common these days, it's kinda the accepted standard. And tested and proven by some top WC experts in the world.

I'm not sure how you will make an efficient heat transfer, and due to the way WC loops work, don't think a uber high flow loop would make any diff. Unless you mean pulling the heat to a fully diff place away from the PC room. Then you would use a heat exchanger, even Koolance makes them. No air channels, all water flow. Not cheap.

maybe your on to something, please explain more. IMOGs idea is valid, but it's pretty much the same as a series loop setup, not two seperate loops.

Conumdrum said:

I'm not sure how you will make an efficient heat transfer, and due to the way WC loops work, don't think a uber high flow loop would make any diff. Unless you mean pulling the heat to a fully diff place away from the PC room. Then you would use a heat exchanger, even Koolance makes them. No air channels, all water flow. Not cheap.

Click to expand...

I'm aware heat exchangers aren't cheap. Neither is my system. I have a custom built DIY chassis with lots of space, as it is for far more than JUST my PC.
When I'm finished my liquid cooling system will be cooling all of the following

-Crosshair V Motherboard (With a full coverage waterblock)

-Vishera 4.0GHz AM3+ CPU (With a 240watt peltier block)

-4x 8GB 2400MHz DDRIII R.A.M.

-Radeon 7990 (possibly 2 in by 2014)

-GTX 460 as a dedicated PPU (It's leftover from my old build. I know it's overkill for a PPU but it's better than leaving it in the closet doing nothing)

-An XBox 360 Corona GPU/CPU

-A Windows 95 computer CPU

-4 SSDs

---Potentially more



Physics dictates that the more waters flows through the heat exchanger the more surface area will be accounted for, thus translating to increase heat transfer. Having fluid flowing in opposite directions would, in theory, add massive thermal transfer to the heat/exchanger.

I will NOT have the two loops flowing into the same reservoir as I want to use Ice Dragon Nano-Fluid for the high pressure loop (It is said to add up to 5degrees Celsius improvement compared to any other fluid, even distilled deionized water)

The hope for this loop is to allow me to have an ultra high flow rate system cool the loop, while a I have a moderate flow rate system actually flow through the waterblocks/heatsinks. This would, hopefully, also decrease the chance of malfunctions, leaks, and increase the life of the system.

There are a LOT of heat exchangers out there, not just the one you linked, though most of them take the form of dual loop passive radiators.


EDIT: The second loops radiators are OUTSIDE the chassis venting to one side on an outtake. The original loop is all intake cooling on the opposite side of the second loop. On top of those I will have additional intake and outtake fans throughout the system to make proper air circulation.

Ice dragon nano fluid is a waste of money. 5C is a lie, and its been tested to be proven as such. Expensive snake oil. Water is the beat fluid. Kill coil or additive to keep things from growing, and that's all you need.

Read the cooling stickies, and it might save you some time. Fluids are one of the basics, and we have around 15 years experience doing this as a community... There's good info around, last ts of mistakes already made so you don't have to.

Just to add to what IMOG said about Ice dragon nanofluid. Skinneelabs tested ICE dragon nanofluid, against distilled water, and it tested slightly worse than distilled water. Although ICE dragon nanofluid is mostly distilled, with some metallic nanoparticles, because of the higher viscosity of nanofluid, the only effect you will see is slightly reduced flow, and slightly worse temps because of the slightly reduced flow, ie by 0.5C or less. Skinnelabs testing is here.

But skinneelabs isnt the only one that debunked the myth of nanoparticles better cooling. Some scientists early on had claimed a positive effect of nanofluids (interestingly only those that had a financial interest made this claim), scientists without a financial interest were unable to see the effect. Then MIT tested 20 different nanofluids from all available sources, using most accurate testing method to date, and showed that none cooled any better than distilled water. Article is here.

Hmmm, well thanks for that heads up on the system, but it still doesn't take away from my desire for a counter flowing dual loop

It's a shame there isn't anything with better thermal conductivity than deionized distilled water for a cooling loop

actually, i believe ammonia (maybe it was paint thinner) has better thermal properties (specific heat?) than water... but it has sort of been decided that NO ONE wants to experiment with a loop filled with that crap. that being said, stick with simple water.

i like the idea you're working with, but it might be overkill and only offer a small benefit apposed to simply adding a nice phat 1080 radiator (or 2). no sense in reinventing the wheel, especially when the wheel already works so well. more radiator&fans > some crazy idea.

dtrunk said:

actually, i believe ammonia (maybe it was paint thinner) has better thermal properties (specific heat?) than water... but it has sort of been decided that NO ONE wants to experiment with a loop filled with that crap. that being said, stick with simple water.

Click to expand...

Ammonia does, however it is caustic and will most likely eat through tubes, additionally you wouldn't be able to get it in a pure concentration it'd be mixed with water (ammonia sold for household use it roughly 5-10% ammonia and the rest water) and of course, any leak would kill you.

Even if paint thinner were better (and it is slightly less lethal), I have to imagine it wouldn't play well with plastic. Most likely would make it brittle.

End of the day, there is nothing that is going to compete with distilled water as far as cost, availability to the average user and of course, the non-lethality of it

Although if you could set some kind of OCing record with ammonia before it killed you..may be worth it

Conumdrum said:

Sounds like a engineers explanation for a series loop.

All needs done is to get the manufacturing to make the parts.

In a CPU/GPU loop, when the CPU is working hard there is extra rad due to the raddage for the GPU also. So the CPU uses the GPU extra raddage to stay cooler than if just a CPU loop.

The same if the GPU is stressed. There is extra raddage due to the CPU being in the loop, so the GPU stays cooler.

That is a naormal CPU/GPU loop which is more than common these days, it's kinda the accepted standard. And tested and proven by some top WC experts in the world.

I'm not sure how you will make an efficient heat transfer, and due to the way WC loops work, don't think a uber high flow loop would make any diff. Unless you mean pulling the heat to a fully diff place away from the PC room. Then you would use a heat exchanger, even Koolance makes them. No air channels, all water flow. Not cheap.

maybe your on to something, please explain more. IMOGs idea is valid, but it's pretty much the same as a series loop setup, not two seperate loops.

Click to expand...

A big +1 on this.

I had dual loop and single loop in the past with the same rig and it overall performed better in a single loop setup.

Rig was
I7 920 @ 4ghz with Apogee GTZ
Rampage III Formula with koolance FC bloc
SLI GTX 580 with EK FC bloc
Rad 1 - TFC360 xchanger (thick rad) with Triebwerk fans
Rad 2 - Swiftech MCR320 with GT's fans
Pump : 4x MCP655 ( i alwais ran 2 pump per loop for safety )

At first, i ran the thick 3x120 rad on the GPU since they output much more heat with 2x MCP655, and another loop with the MCR320 + 2 pump for the I7 and mobo. Temp were fine but i decided to try with everything on the same loop with 2 rad and 2 pump. CPU temp remained about the same but GPU temp dropped quite a bit since while gaming, the CPU dont output that much heat so the whole 6x120mm rad of cooling was much better at cooling the whole rig than everything seperate.

If you really want to experiment... you don't really need a "heat exchanger" in the traditional sense, you're using the same cooling medium... use a reservoir and mix. have your blocks on one side of the loop, all your radiators on the other side of the loop.

Now if you really have a stiffy for a plate and frame heat exchanger like the Koolance, you'll want a bigger temperature differential between hot and cold sides... TECs or Phase assisted maybe?

The 'real 'pro has spoken. Pay attention to his input. You ever visited Skinneelabs.com?