DOES SIZE REALLY MATTER?? - Tubing Size Showdown

[jakkuh_t]

The question of which tubing size to buy has long plagued the watercooling community, but finally the case is closed.

 

 

Buy Primochill Tubing:
On Amazon: http://geni.us/MR4jJ
On Newegg: http://geni.us/JsLn

[Daniel644]

well basic physics says beyond the diameter of the inside of the fittings or the tightest point in the loop it doesn't matter, you can't shove any more flow through then what that area allows through so bigger tubing would only equate to a higher total volume of water in the loop, by a small amount. The head pressure and fitting diameters are the controlling factors in this not tube size, unless tube size ends up being the smallest point in the loop.

[johnnydoe9]

Seriously though when is Luke hosting another video?

[Eulion]

I was about to upgrade my cooling cause I thought it would be better with bigger tubes. I guess I was wrong.
And thanks guys for helping me save a bit of cash.

[Arihant]

2 hours ago, jakkuh_t said:

The question of which tubing size to buy has long plagued the watercooling community, but finally the case is closed.

 

 

Buy Primochill Tubing:
On Amazon: http://geni.us/MR4jJ
On Newegg: http://geni.us/JsLn

Guys why waste time discussing the tube size when you can hire an engineer

[Marinatall_Ironside]

1 minute ago, Arihant said:

Guys why waste time discussing the tube size when you can hire an engineer

Because engineers, like other professionals in their respective fields relevant to this topic only care about theories, not results or practical application.

[Enderman]

2 hours ago, Daniel644 said:

well basic physics says beyond the diameter of the inside of the fittings or the tightest point in the loop it doesn't matter, you can't shove any more flow through then what that area allows through so bigger tubing would only equate to a higher total volume of water in the loop, by a small amount. The head pressure and fitting diameters are the controlling factors in this not tube size, unless tube size ends up being the smallest point in the loop.

Kinetic losses of fluid flow is NOT just the smallest point in a loop, it depends on the sum of all minor and major (frictional) losses.

Larger tubing equates to less fluid velocity, and therefore less energy loss.

https://en.wikipedia.org/wiki/Minor_losses_in_pipe_flow

Notice the velocity squared relationship.

The difference in diameters is so small, and the pumps used in watercooling have so much power, that the difference isn't practically measurable, as seen in the video.

 

There IS however, a difference.

Having a larger diameter pipe and lower fluid velocity is more efficient for a given flow rate, which is why stuff like water or oil pipes often have huge diameters and low velocity.

 

Basically, for PC watercooling, pick the tubing you think looks best and can do the bends you want.

With a good pump there is no measurable difference.

For stuff other than PC watercooling, you may need to take into account pipe diameter.

[Luscious]

Wow SMH. You guys really do need to go back and either do your homework or consult with a water cooling expert. It's not the tubing size or even wall thickness that comes into play, but rather the VOLUME of coolant your loop can hold - the more the better.

 

Regardless of your tubing length, a 13mm ID tube vs a 10mm ID tube will pass around 70% more water through. Multiply that out by how long your tubing actually is, and you could be looking at anywhere from 8oz to 32oz more coolant in your loop.

 

It's a very simple equation. More volume = more coolant to absorb heat = lower delta temperatures. I have close to a gallon of the stuff in my rig cooling five massive blocks. My delta stays between 15 and 18C remarkable constantly.

 

Pump flow rate specifies how much coolant circulates over time, but with a bigger diameter tubing, that equates to better flow rate at a lower pressure, rather than having a restricted flow rate and work the same pump at a higher pressure. That higher pressure will then compete with your pumps maximum head pressure limit - something that will also be affected by how many 90 degree fittings you have in your system, the number of blocks/radiators and the placement of your pump relative to the reservoir. And since gravity is your best friend, the ideal place to put your pump is to mount it near the bottom of your loop and place the reservoir directly feeding it up top.

 

For very complex loops you can also use two pumps in series. This has the added benefit of keeping your system running if one of the pumps dies, avoiding overheating and shutdown. The need to actually run TWO SEPARATE LOOPS (one for the GPU and one for the CPU) is pure money waste and has no performance benefit whatsoever.

 

I use 7/16 ID 5/8 OD tubing in my system with EK compression fittings. These measure 22mm across on the outside. Any larger and the compression fittings required won't clear the fan mounts on a standard radiator unless you use a fitting extension, which would just look FUGLY. That's what rules out the use of the biggest size 1/2 ID 3/4 OD tubing, at least in my case.

 

Hard line fittings may have a different outer dimension, but you still need to keep in mind fan clearance on radiators when attaching your fittings on the same side.

 

So yeah, SIZE DOES MATTER

[Enderman]

7 minutes ago, Luscious said:

Wow SMH. You guys really do need to go back and either do your homework or consult with a water cooling expert. It's not the tubing size or even wall thickness that comes into play, but rather the VOLUME of coolant your loop can hold - the more the better.

 

The point of watercooling is not to use the water to store heat............

Heat is dissipated by the radiators.

Having more water simply means it takes longer to heat up, it does NOT improve cooling.

 

The reason they run tests for more than 10 minutes is to let the loop come to equilibrium, aka heat dissipation matches the heat input.

This equilibrium temperature will be the same regardless of how much fluid is in the loop.

If you put a giant 2L reservoir maybe it will take 100 minutes instead of 10 but nobody cares about that, what matters is the continuous heat dissipation not the warm up time.

 

The volume of your loop does not matter.

Having a bigger reservoir does not improve cooling.

I think you're the one who needs to do their homework on watercooling again.

[Motifator]

This guy tries to sound like he knows what he is talking about but he doesn't.

How big of a res / how much water you have in your loop equates to VERY LITTLE performance difference. The only thing water does there is to CARRY heat, just because you have more water or coolant doesn't mean you'll get better temps. If you want the absolute best temps, you should be looking at BETTER rads / blocks and higher head pressure pump. You can have the best loop in the world with a cheap small plastic res. For all the loop cares, performance won't change between that and a big quality res.

[DildorTheDecent]

10/16 = no kink

10/13 = kink ( ͡° ͜ʖ ͡°)

 

I know what I'm going for.

[Daniel644]

50 minutes ago, Luscious said:

Wow SMH. You guys really do need to go back and either do your homework or consult with a water cooling expert. It's not the tubing size or even wall thickness that comes into play, but rather the VOLUME of coolant your loop can hold - the more the better.

 

Regardless of your tubing length, a 13mm ID tube vs a 10mm ID tube will pass around 70% more water through. Multiply that out by how long your tubing actually is, and you could be looking at anywhere from 8oz to 32oz more coolant in your loop.

 

It's a very simple equation. More volume = more coolant to absorb heat = lower delta temperatures. I have close to a gallon of the stuff in my rig cooling five massive blocks. My delta stays between 15 and 18C remarkable constantly.

 

all the coolant in the world isn't going to matter if you don't have the radiator space to dissipate the heat the coolant picks up the tubes aren't radiators, more water will only prolong how long it takes to reach the peak temperature that the radiator will hold the equipment at.

[Egg-Roll]

4 hours ago, Jurunce said:

Because engineers, like other professionals in their respective fields relevant to this topic only care about theories, not results or practical application.

Stop watching the Big Bang Theory... Engineers in fields like this can't soly go based off of theories, because if they did then theoretically a 1mm thick tub made up of special composites could hold 10,000 PSI, but would it be logical in any mass production product?

 

Your home was touched by a engineer at some point and if you live in a 5+ story building if I were you I'd rather logical facts over theories.

 

Any theories they have need to be tested in a lab before it can be put into practice.

 

4 hours ago, Arihant said:

Guys why waste time discussing the tube size when you can hire an engineer

Are you buying? If not then no.

 

6 hours ago, jakkuh_t said:

The question of which tubing size to buy has long plagued the watercooling community, but finally the case is closed.

The better question is why didn't you guys get Linus to surf on coolant with shades using the image at 1:55? Or too baitclicky

[Arika]

4 hours ago, johnnydoe9 said:

Seriously though when is Luke hosting another video?

he's not obligated to (also to be honest, he also shouldn't) because he doesn't work for LTT anymore.

[choLOL]

There are a lot of factors affecting fluid flow in pipes, not just diameter. I may lay out some equations for illustration purposes only, not gonna solve unnecessary problems. These formulae come from very basic chemical engineering fluid mechanics; bachelor's degree level only, no MS or PhD level stuff here. TL;DR at bottom.

 

And please, inb4 someone says these are only ideas and concepts which do not reflect real world applications such as computer water cooling. It's very irritating when someone says actual math and physics is total shit compared to "real world applications". B hows your pc and water heating up without math and physics bruh wtf. We work with real world scenarios, using math and physics to somewhat predict what will happen to real world systems.

 

1.) Mechanical energy balance, friction losses, and pressure drop

The MEB (attached picture at the very bottom) basically means:

[Total mechanical energy loss of the fluid due to frictional forces] +

[Pressure drop of the fluid from one point to another (such as right after the pump til the end of the loop)] +

[Total change in kinetic energy, aka change in flow rate] +

[Total change in potential energy, aka height differences in the loop, sometimes irrelevant]

is equal to

[Shaft work, aka work performed by the pump]

 

Flow rate in terms of mass/kilograms of fluid per second, kg/s, is constant throughout a loop when there there are no parallel branches. Simple explanation is there's nowhere else the mass will go, it'll go through the loop in a constant mass rate if the fluid is liquid and therefore incompressible.

 

The part where the debates usually occur is in the friction losses, resulting in pressure drops along the loop. Friction losses occur due to (1) sudden expansion of flow area such as when fluid exits a pipe and enters the reservoir, (2) pipe bends, (3) flow restrictive parts, (4) surface roughness of the pipe.

 

The pipe diameter does indeed affect fluid flow. It's a really big factor. There's the reynolds number, fanning's friction, churchill equation, among others. But the pipe diameter will only affect change in temps if fluid flow is poor.

 

There's a whole lot to be said about friction loss, but as long as the pump can supply greater head than the friction loss, it's all good. Go find pumps that give good head, iykwim, and pumps that won't die quickly.

 

 

2.) Amount of water in the loop

The amount of water in the loop also helps with temps. The formula for heat is as follows:

Q = m Cp (T2-T1),

or (heat) = (mass) x (specific heat) x (change in temps).

More water = longer for the loop fluid to change temp.

For example, take a cpu which outputs 65w, or 65 joules/sec. Take one second, only 65 joules of heat is supplied. Ten liters of water will only have a really small temperature change due to 65J, but the temps of 250mL of water will change much more abruptly. If a steady amount of heat is supplied for a prolonged amount of time, both of them can stabilize at the same temperature. The difference is the amount of time for the water and chips to abruptly change temps.

 

 

 

 

I really don't know where I'm going with this, it's more of a response to others' claims when it comes to liquid cooling, in this thread and in all threads, including other youtubers and their videos about water cooling. There are a lot more formulae, especially for friction losses, but I don't see the point of posting them here.

 

TL;DR: It's very simple when you use the corresponding formulae.

More water = good

Less friction = good

Diameter --- not as important as good head

 

If anyone wants to learn more about fluid flow, there are a lot of good youtube videos simplifying fluid mechanics. I used them back in the uni. Or read up on the overall energy balance, mechanical energy balance, reynolds number, and maybe fanning's friction equation, then go on to more advanced stuff.

 

.

[Arika]

1 hour ago, choLOL said:

And please, inb4 someone says these are only ideas and concepts which do not reflect real world applications such as computer water cooling. It's very irritating when someone says actual math and physics is total shit compared to "real world applications". B hows your pc and water heating up without math and physics bruh wtf. We work with real world scenarios, using math and physics to somewhat predict what will happen to real world systems.

the math is nice and all, except the "real world scenario" that was tested showed that none of what you just said mattered. the temperatures were within margin of error from each other

[choLOL]

3 hours ago, Sierra Fox said:

the math is nice and all, except the "real world scenario" that was tested showed that none of what you just said mattered. the temperatures were within margin of error from each other

Well if you actually read my tldr, you would understand, no? Also, I was talking about fluid flow, which is a totally separate topic from the heat transfer aspect of water  cooling. I don't think I stated anywhere that the fluid flow depending on diameter heavily impacts temps. I even stated that having tons of water would stabilize at the same temps as having less water, given that everything else stays the same.

 

I'm confused. Wat

[Canada EH]

On 6/2/2018 at 12:11 PM, jakkuh_t said:

has long plagued the watercooling community

has it really?

 

come on now be honest.

[GeorgeDouj88]

On 6/2/2018 at 3:32 PM, johnnydoe9 said:

Seriously though when is Luke hosting another video?

Are you serious? He left linus tech tips, he works at floatplane which is for ltt and in the same building but he doesn't do videos anymore