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

#### theoneandonlymrk

would slowing down the flow through a radiator by half increase efficiency.

I ask because that is what I intend to do, split the water loop after the heat loads into two 360 radiators ,one 60mm one slim 25mm so as to slow the flow down(the fans would be using the same profile), my thinking being that my high flow rate pushes it through the radiators fast and slowing it down would give it more time in the radiator to cool.
single loop beyond this though ie to the gpu and cpu etc.

#### agent_x007

If you can't cool your water to near ambient on 360mm long 60mm thick radiator - you are probably doing something wrong... Just pump more air through radiator.

#### Vario

would slowing down the flow through a radiator by half increase efficiency.

I ask because that is what I intend to do, split the water loop after the heat loads into two 360 radiators ,one 60mm one slim 25mm so as to slow the flow down(the fans would be using the same profile), my thinking being that my high flow rate pushes it through the radiators fast and slowing it down would give it more time in the radiator to cool.
single loop beyond this though ie to the gpu and cpu etc.
No, that is incorrect. Equation for heat transfer indicates that increased flow rate increases heat transfer.

https://www.overclockers.com/water-cooling-flow-rate-and-heat-transfer/
There is an elementary equation from basic thermodynamics that states that the rate of heat transfer (Q) equals the mass flow rate (M) times a Constant (the specific heat of water) times the Delta T (fluid temp out minus fluid temp in):

Q = M x C x Delta T
If you increase flow rate with the same pump, your temperatures will trend in the direction of goodness. If you increase flow rate by going to a bigger pump, you will reach a trade off somewhere where the pump starts putting too much energy into the system and temperatures will start increasing.

#### theoneandonlymrk

No, that is incorrect. Equation for heat transfer indicates that increased flow rate increases heat transfer.

https://www.overclockers.com/water-cooling-flow-rate-and-heat-transfer/
I am specifically trying to optimize my main system, its flat out 24/7 in my room so noise has to be expected but its managed so the dual 18watt laing ddc pumps i run are set to a steady 82% solid upto 75 degrees then they enter panic mode with the rest of the pc but all my fans also run within a tight window since i know what loads its doing my fans are stuck between 78-88% and all is well ,but this is relatively stock, just PBO bossting it to 4ghz automatically and for more and need to dump some heat.
that article describes it how I expect and know the CPU part to be ,higher flow to a point does good hence my 82% ,its a tested top of the bell curve figure, however heat raidation water to metal to air might pan out a bit different , should i just blame sleep depprevasion and skulk off?

Last edited:

#### NdMk2o1o

You shouldn't even be hitting 75c on a custom 360 loop as I can run my 1st gen ryzen 1600 at 4ghz under 70c on a relatively cheap raijintek leto Pro, are you sure your, wc setup is running as it should?

#### theoneandonlymrk

You shouldn't even be hitting 75c on a custom 360 loop as I can run my 1st gen ryzen 1600 at 4ghz under 70c on a relatively cheap raijintek leto Pro, are you sure your, wc setup is running as it should?
not entirely i'd have to say or i would'nt still be messing with it but its a work in progress and still quite new to me, ive only rebuilt it once , but also, my one is at 70 degrees max not 75 and as i said thats both the cpu and gpu 100% loaded 24/7 after 24 hours and stable obviously and with the pumps at 82% and fans hitting about 88% constant.

also PBO uses the cooling you have to max the clocks against higher then stock wattage so it will run to its wattage limit more then thermal in my case.

err and no worries guys im not going to argue against thermodynamics ive put it down to a brain fart, it happens but i still learned a positive in that my pumps are probably putting more heat in then i think and i can counter that a bit , win.

Last edited:

#### Vario

What type of fans? Is it possible the wrong kind of fans for the type of radiator you are running? Some fans move a lot of air but only if unobstructed and therefore work very poorly against a radiator.

#### Gasaraki

I really depends on your scenario. Faster pump 'generally' is better cooling but it depends on what type of radiator you have, the amount of blocks you are cooling, the resistance of the flow in your loop, etc. Most of the time slowing down the pump does not affect the cooling performance much if everything is set up correctly. If you are running slow fans then maybe running your pump at a medium speed might help.

In your case, you have so many radiators that it doesn't matter what speed you are running the pumps, you just need to make sure the pump is good enough to push the resistance of all those radiators and blocks.

Last edited:

#### phanbuey

Pump speed is a bit like ram size - once there's 'enough' flow, more doesn't really add anything. Slowing it down past a certain point would definitely cause everything in the loop to start heating up, including the radiator. So yes, in theory, pumping hotter fluid through your radiator would increase 'efficiency'.

I usually run my pumps at half speed since it doesnt make any temp difference for a simple loop and quiets the machine substantially.

#### theoneandonlymrk

What type of fans? Is it possible the wrong kind of fans for the type of radiator you are running? Some fans move a lot of air but only if unobstructed and therefore work very poorly against a radiator.

all the fans were made for radiators, and @NdMk2o1o ,are you sure were talking about the same cpu temp even , i can name 4 that all read different ,i mention this because Ryzens max is 85 so 70 actual in the core temp is'nt bad flat out, my cpu on the whole sits at 60 degrees and the loop at 36min cold point 46 hottest point max atm, my fans are getting a bit of an upgrade in the process anyway.

just i am draining it and replumbing anyway plus moving stuff about a bit , this was a thrown together build run since i had no new pipe when i built it so got thinking along the can i improve it point.

I am going to improve it regardless of the Op im fitting 3 "in" fans that are not there atm and a slight rejig of order so its not a no hope issue.

sorry for the misquote Vario

Last edited:

#### infrared

##### Moderator
Staff member
I'd switch out the ddc pumps for D5 pump if they're struggling and getting too hot.

As the others have said, up to a point you benefit from flow rate, going too slow will definitely hurt cooling performance though.

Edited for terrible grammar!

Last edited:

#### Steevo

not entirely i'd have to say or i would'nt still be messing with it but its a work in progress and still quite new to me, ive only rebuilt it once , but also, my one is at 70 degrees max not 75 and as i said thats both the cpu and gpu 100% loaded 24/7 after 24 hours and stable obviously and with the pumps at 82% and fans hitting about 88% constant.

also PBO uses the cooling you have to max the clocks against higher then stock wattage so it will run to its wattage limit more then thermal in my case.

err and no worries guys im not going to argue against thermodynamics ive put it down to a brain fart, it happens but i still learned a positive in that my pumps are probably putting more heat in then i think and i can counter that a bit , win.

I use a single 360 closed loop to cool my CPU first then my GPU that is pumping out massive heat and it only reaches 55C on the GPU running F@H 24/7

#### Dinnercore

As the pump question has been answered now, IÂ´d say basic tips to optimize radiator setups is care for enough airflow with fans build for this cause, consider using push/pull on the thick radiator, try to avoid any airflow obstructions, if one rad is intake and one exhaust, try to match the fans speed so that there is no immense negative or positive pressure inside the case to avoid turbulent air or air leaking elsewhere.
Is the ambient temp high in this room? I found it helpful to leave doors or a window open in a room with a 24/7 running machine that will eventually heat a whole room by some amount.

#### Vario

What is the model # of fans? Waterblock clean of algae? Any mixed metals in loop? Just some things to think about if you are dissatisfied with the performance.

#### phanbuey

Also if you can use the coldest possible air for your radiators, having them as intakes, or having them use fresh air straight from an intake makes a huge difference.

This is especially true for quiet setups - I've had a push pull 120mm intake edge out a 240mm exhaust AIO due to the fact the 240 was cooling using the hot air from the case which was already +5-10 C over ambient.

#### phill

Just a question I take it all of this is a single loop? Have you considered putting the CPU in one loop and the GPUs in another?? (Went by your system specs, so I hope this is right )

#### hat

##### Enthusiast
Interesting... I would have thought the higher the flow rate, the better. Guess not?

#### phanbuey

I would think you would have to have special equipment to get the flow fast enough to where absorption is an issue - all the tests i remember seeing from Martin's Liquid lab showed an inverted curve increasing dissipation with flow although at increasingly tiny amounts. I'm sure at some point there is a friction component once you're blasting the coolant at ludicrous speed.

#### Steevo

Interesting... I would have thought the higher the flow rate, the better. Guess not?

Depends on the pump, loop size, blocks and much else. Generally higher flow is better than lower for all situations unless you are trying to force it through a too restrictive block or something, then the pump will add heat to the block. To double a flow through a given orifice you have to quadruple the pressure.

#### hat

##### Enthusiast
I would think you would have to have special equipment to get the flow fast enough to where absorption is an issue - all the tests i remember seeing from Martin's Liquid lab showed an inverted curve increasing dissipation with flow although at increasingly tiny amounts. I'm sure at some point there is a friction component once you're blasting the coolant at ludicrous speed.
You are correct!

Depends on the pump, loop size, blocks and much else. Generally higher flow is better than lower for all situations unless you are trying to force it through a too restrictive block or something, then the pump will add heat to the block. To double a flow through a given orifice you have to quadruple the pressure.
The only time I've seen flow rate add heat to anything was back when I did industrial cleaning. We had water pumps mounted on trucks which pushed water at a very high pressure (up to 10k PSI usually, more on other systems). The barrel on the gun at the business end of that thing would get quite warm, even hot to the touch. Fortunately, I don't think anyone is cooling their PC with waterblasting equipment...

That's a really extreme example, but I don't see any normal water cooling pump adding heat to anything...

#### thebluebumblebee

Jayztwocents: "Does pump speed matter when watercooling?"

#### cdawall

##### where the hell are my stars
I can cool my 2990wx passively with my current loop. Pump speed doesn't change Temps really once you bit equalibreium. I would try and find out why there are temp issues with the loop if you are getting a bit hot.

Could be anything from an awkward bubble trapped in the rad, block mounted to tight or even a clogged component.

#### hat

##### Enthusiast
What do you mean passively? No pump? How does that work?

#### Jetster

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