• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Would slowing down the flow through a radiator by half increase efficiency

What do you mean passively? No pump? How does that work? :fear:

Passively meaning no fans spinning :) Just uses the ambient air temp to cool :)
 
First let me say I know nothing about custom loops. That being said there is a reason cars have a thermostat. Besides just to warm the car up. It also regulates the speed of the water allowing it time in the rad to be cooled. So yes I would guess you can pump the water too fast

But, cars don't have an adjustable 'fan' whilest driving. And since an engine needs to remain at a certain opperating speed, I'd rather think this flowadjustment has more to do with that..

.. though I don't know for sure.

Having said that, I'm wondering, isn't it a good idea to replace the water with car coolant? Might seem like a dumbass idea, but.. car coolant does have some noticeable heat conductivity improvement over water. Though I'm definitely not sure one can 'just go ahead' with that idea. This is purely an hypothesis, and I have no PoC, nor any idea what difference there is between chemical interactions on the car cooling tubes and watercooling tubes, when you switch water with coolant!

It's just an hypothetical thing.
 
But, cars don't have an adjustable 'fan' whilest driving. And since an engine needs to remain at a certain opperating speed, I'd rather think this flowadjustment has more to do with that..

.. though I don't know for sure.

Having said that, I'm wondering, isn't it a good idea to replace the water with car coolant? Might seem like a dumbass idea, but.. car coolant does have some noticeable heat conductivity improvement over water. Though I'm definitely not sure one can 'just go ahead' with that idea. This is purely an hypothesis, and I have no PoC, nor any idea what difference there is between chemical interactions on the car cooling tubes and watercooling tubes, when you switch water with coolant!

It's just an hypothetical thing.

The fan is regulated in a car with a clutch or on electric fans will start and stop. Antifreeze helps keep corrosion away and does aid in cooling in warm weather, better than straight water. Water is corrosive. ... Yes, you can run straight water, but your coolant boiling point will be lower
 
The fan is regulated in a car with a clutch or on electric fans will start and stop. Antifreeze helps keep corrosion away and does aid in cooling in warm weather, better than straight water. Water is corrosive. ... Yes, you can run straight water, but your coolant boiling point will be lower

Sorry, I was referring to the wind, or 'fan' coming in through the front, as a car moves. Not the actual radiatorfan - sorry I only drove old crap cars (€300,-), in fact my current Starlet( :rockout:the good model, with injection and sturdy side mirrors hard as brick, old, but still pumping to that 170km/h like a beast) might actually be the first to have a working radiatorfan :fear:- my bad. I ment no offence, mate, sorry if it came over so. :(
 
Sorry guys I have not got heat issues but thanks, im technically trying to reduce noise here , i can easily up the pump speed and fans but thats the opposite of what im trying to achieve.

As I said in the Op it's on 24/7 full load on a monoblocked cpu and full waterblocked vega 64 so plenty of heat im just trying to optimise it to get the noise lower, it's a bedroom pc too.

Also were not talking whole loop, the heat sources would be full flow then a split to two rads to reduce the flow while at the same time the fans would remain at 78% , my annoying noise threshold apparently.

And so far I have decided thermodynamics wins but since I have teo rads I can pass slightly cooler water to the system by going res, pump, rad ,load, rad ,res instead of as it is now res, pump, load , rad,rad.
 
I have to repeat this every 6 months or so when people come with water cooling questions. It doesn't matter what order you put your parts. Within a few minutes the water inside your loop reaches an equilibrium and ordering ceases to matter, since all the liquid in a loop becomes the same temperature.
 
I have to repeat this every 6 months or so when people come with water cooling questions. It doesn't matter what order you put your parts. Within a few minutes the water inside your loop reaches an equilibrium and ordering ceases to matter, since all the liquid in a loop becomes the same temperature.
And I recall telling you that very small margins and 2-4° less heat at the coldplate of a part can enable Overclocks you can't with that mindset ,one that doesn't account for multiple radiators or both getting cold air or not.
At times this pc games or benches , pliability is key.
And I ran three rads in a fair few configurations with the same case and load and found for example the last gpu in line via your plan wouldn't match the clock of the first, move a 120 rad from the end of the loop to between the cards and they matched top clocks Rx480Wb(EK) @1366/2100 24/7 loaded not a bench , without that in-between rad it wouldn't bench at 1366 nevermind run 24/7.
 
Sorry guys I have not got heat issues but thanks, im technically trying to reduce noise here , i can easily up the pump speed and fans but thats the opposite of what im trying to achieve.

As I said in the Op it's on 24/7 full load on a monoblocked cpu and full waterblocked vega 64 so plenty of heat im just trying to optimise it to get the noise lower, it's a bedroom pc too.

Also were not talking whole loop, the heat sources would be full flow then a split to two rads to reduce the flow while at the same time the fans would remain at 78% , my annoying noise threshold apparently.

And so far I have decided thermodynamics wins but since I have teo rads I can pass slightly cooler water to the system by going res, pump, rad ,load, rad ,res instead of as it is now res, pump, load , rad,rad.

I have no idea how watercooling works, but you could try to use as long as possible and thin hoses, in combination with an exhaust case fan. Thin hoses might increase your mobility, letting you guide the hoses allong anything that is cooler, along the way.

Using the cooling you have as efficiently as possible, and double dipping as much as possible.

I dont know if you can just use hoses and mod as you please with watercooling, but you could even just take some heatsinks you might have lying around, drill appropriate holes for your hoses, and guide them through there.

As long as your case temp stays cool, anything heatsinky will have the potential to passively buff the heat you lose along the way.

If your trully a nutter, I'd say: man can do much with copper pipelines, a blowtorch and a cognitively drawn design architecture.
As far as plastic hozes go, the bendy ones usually feel damn isolating. Dunno if you're handy with a soldering iron an a blowtorch, though?

I have to repeat this every 6 months or so when people come with water cooling questions. It doesn't matter what order you put your parts. Within a few minutes the water inside your loop reaches an equilibrium and ordering ceases to matter, since all the liquid in a loop becomes the same temperature.

That doesn't make any sense. Liquid leaving the CPU, will be hotter than that same Liquid entering whatever comes after. Or GPU. Or whatever hot that it is cooling.

Seeing CPU's are generally out run by GPU's, in my experience and world, i'd stick CPU before GPU, with or without a 'rad'(correct me if I'm wrong, this is to be the radiator that cools the liquid, right? ) in between. Simply cause hot -> cool takes longer than very hot -> cool, vice versa. Thats passive cooling you're throwing away.

You say "keep the change" alot when shopping? I do to, but thats cause I'm usually not making any sense in public places with lots of stimuli, and no clear goal.
 
Last edited:
halving flow rate would not increase efficiency. it would balance out the same, but you can have a flow rate which is too low
Also loop order does not matter as the entire system is a closed circuit it will reach an equilibrium regardless of where you have the radiators. More radiators can affect the temperature of the water but where they are makes no difference to the loop.
The onlu thing that matters is the pump needs to be Below the reservoir and the 1st thing after the reservoir so it gets no issues receiving water.
Anything after that makes no difference regardless of how counter intuitive it may seem.

You can argue that it will be a degree or 2 colder But that's only before equilibrium is reached, after that then the system will maintain the temps, and the temp of the water will remain a constant. (its a closed loop/circuit) The radiators can only dissipate the heat they are able too the rest of the heat will stay in the system and the water will balance it out.
 
that's only before equilibrium is reached, after that then the system will maintain the temps, and the temp of the water will remain a constant.
Not true. The water temperature drops while passing inside the radiator. Inlet temp is always higher than the outlet, and the delta between them, combined with the water flow, represents the heat rejection to the ambiant air.
Also, the same applies to the water blocks. Inlet temperature will be lower than the outlet, because of heat extracted from the component. Again, this is in relation with water flow.

If you have a very high water flow, the delta temp will be very small. But then you are wasting energy in pumping losses.
 
First let me say I know nothing about custom loops. That being said there is a reason cars have a thermostat. Besides just to warm the car up. It also regulates the speed of the water allowing it time in the rad to be cooled. So yes I would guess you can pump the water too fast
I don't think the analogy to PC watercooling is correct. You want to bring the car's engine up to operating temperature quickly so that wear and tear on the mechanical surfaces such as bearings and cam lobes is reduced. Once at operating temperature you want to keep it there so it is not overheating. With a PC you want it to run cool at all times. The thermostat's purpose is maintaining the engine at the engineer specified operating temperature.
 
Last edited:
The fan is regulated in a car with a clutch or on electric fans will start and stop. Antifreeze helps keep corrosion away and does aid in cooling in warm weather, better than straight water. Water is corrosive. ... Yes, you can run straight water, but your coolant boiling point will be lower

Coolant boiling point is higher than water.
 
I have some concerns:
First, that you might have the wrong fans, or that you're trying to run them too slowly to push air through a 60 mm rad.
Secondly, if you split the flow and run the rads in parallel, water will flow through the path of least resistance so virtually all of the flow could end up going through only one of the rads, unless you got flow indicators and a valve to balance the flow. I'd think you'd have to go to 2 loops or run them in series. You also list 2 DDC pumps. If those can be run in parallel, then you could connect the output of each to a rad.
Third, that the case you have is an OC'ing case that was not designed for low noise, but then again, I have never seen a case that is quiet with a high heat load.
 
Regarding the question about slowing the liquid down in the radiator

The myth is stated as either:
  1. Coolant can be pumped too fast through the engine for it to absorb enough heat, or
  2. Coolant can be pumped too fast through the radiator for it to cool properly, or
  3. Cooling can be improved by slowing the flow of coolant through the radiator so it cools more completely.
NONE of these is true. The simple truth is that higher coolant flow will ALWAYS result in higher heat transfer and improved cooling system performance.

TL: DR (But a lot of good info related to cooling): http://www.pirate4x4.com/tech/billavista/Cooling/
 
As mentioned halving the flow rate will DECREASE your overall efficiency. Can you post up some pictures of your case so we can get a better handle on your watercooling setup?
 
I have some concerns:
First, that you might have the wrong fans, or that you're trying to run them too slowly to push air through a 60 mm rad.
Secondly, if you split the flow and run the rads in parallel, water will flow through the path of least resistance so virtually all of the flow could end up going through only one of the rads, unless you got flow indicators and a valve to balance the flow. I'd think you'd have to go to 2 loops or run them in series. You also list 2 DDC pumps. If those can be run in parallel, then you could connect the output of each to a rad.
Third, that the case you have is an OC'ing case that was not designed for low noise, but then again, I have never seen a case that is quiet with a high heat load.
Don't sweat it bro some of my fans are crap ish but not terrible.

And I didn't split it as i have said thermodynamics laws wins, physics wins.
I had my hands up claiming a brain fart posts ago:)

Your post is my favourite since you had concerns but actually answered the Op in a way i understand and agreed with, you would have changed my mind:) and then voiced your concern most thought I had an issue and are trying to fix it.

To others my loops cooler and quiter now i have a rad after pump and before gpu, it benches with higher clocks too even when soaked prior but no

Splitting a loop into two rads wont increase efficiency.

The laws of thermodynamics forbade it.

Presently my pc is setup thus reservoir ,two laing ddc with aquacomp dual head in series for head and redundancy ,360 slim Ek rad , vega64 , R2600x, 360 (60mm) rad , reservoir, pretty traditional.
 
What is the brand and model of radiator fan? Do you have any instance of mixed metals in the loop or any possibility of algae growth?
 
The fan is regulated in a car with a clutch or on electric fans will start and stop. Antifreeze helps keep corrosion away and does aid in cooling in warm weather, better than straight water. Water is corrosive. ... Yes, you can run straight water, but your coolant boiling point will be lower


You are wrong, glycol has a lower heat transfer rating than water despite increasing it's specific density.

Running pure water vs a 50 percent solution of water and ethelene or propolene glycol will result in a 20-25 percent reduction in heat transfer. My new house is geothermal and uses water and methanol as it's superior to glycols pumping and hest transfer abilities. I have charts showing this.

Water wetter or other coolant additives reduce the surface tension of plain water and allow it to have a greater themal conduction. I use water, 15% glycol, water wetter, and my system.has been running essentially 24/7 for 7ish years.
 
You are wrong, glycol has a lower heat transfer rating than water despite increasing it's specific density.

Running pure water vs a 50 percent solution of water and ethelene or propolene glycol will result in a 20-25 percent reduction in heat transfer. My new house is geothermal and uses water and methanol as it's superior to glycols pumping and hest transfer abilities. I have charts showing this.

Water wetter or other coolant additives reduce the surface tension of plain water and allow it to have a greater themal conduction. I use water, 15% glycol, water wetter, and my system.has been running essentially 24/7 for 7ish years.

Water wetter you say? Gonna have to give that a try.

Is there anything that will reduce evap beyond the propylene glycol? Im putting in an Alphacool Eisbear this weekend and filling it is a huge pain in the @$$. It's kind of an AIOish watercooling system and Im trying to figure out a way to limit fillings.

Should i just go straight glycol?
 
Water wetter you say? Gonna have to give that a try.

Is there anything that will reduce evap beyond the propylene glycol? Im putting in an Alphacool Eisbear this weekend and filling it is a huge pain in the @$$. It's kind of an AIOish watercooling system and Im trying to figure out a way to limit fillings.

Should i just go straight glycol?


My loop is closed and I generally only add a few ml of water a year, I wouldnt run straigt glycol as its viscosity is too high to be pumped easily, and it holds way less heat than plain water. I suggest Distilled or deionised water, water wetter or another good surfactant, and 10-25% of generic car antifreeze ( premixed 50/50, or half of a concentrate) unless you want a colored coolant. I have only changed my coolant a few times and each time the system is pristine. Im going to try OAT coolant next as it has a higher thermal capacity than glycol according to manufacturing specs.
 
My loop is closed and I generally only add a few ml of water a year, I wouldnt run straigt glycol as its viscosity is too high to be pumped easily, and it holds way less heat than plain water. I suggest Distilled or deionised water, water wetter or another good surfactant, and 10-25% of generic car antifreeze ( premixed 50/50, or half of a concentrate) unless you want a colored coolant. I have only changed my coolant a few times and each time the system is pristine. Im going to try OAT coolant next as it has a higher thermal capacity than glycol according to manufacturing specs.
I am using EK kryotherm premix blood red :) .
It's been crunching and folding all day my hottest water temp is 36° c my socket is 46 gpu 36/ hotspot 46 cpu cores are 68 internally and 59 as a whole ,ill probably reduce fan speed a bit they're running at 1200 rpm atm , they're 3xcorsair LL120, 3x xigimatic 120 1300rpm and 3x bitfenix 120 fans

So it's better ,but not miles better my cores Were sat at 75 before so big smiles.
Just wondering about tightening the monoblock some more but it's cranked as much as possible without messing with standoffs (the monoblocks)mod style.
 
You are wrong, glycol has a lower heat transfer rating than water despite increasing it's specific density.

Running pure water vs a 50 percent solution of water and ethelene or propolene glycol will result in a 20-25 percent reduction in heat transfer. My new house is geothermal and uses water and methanol as it's superior to glycols pumping and hest transfer abilities. I have charts showing this.

Water wetter or other coolant additives reduce the surface tension of plain water and allow it to have a greater themal conduction. I use water, 15% glycol, water wetter, and my system.has been running essentially 24/7 for 7ish years.
No, Google it, its a quote

Coolant boiling point is higher than water.

That's what it said "Water is corrosive. ... Yes, you can run straight water, but your coolant boiling point will be lower" Which is why water is not wetter. It turns to steam first
 
Last edited:
Back
Top