Overclock.net › Forums › Cooling › Water Cooling › A few words about pumps and flow rate.
New Posts  All Forums:Forum Nav:

A few words about pumps and flow rate. - Page 3

post #21 of 74
Quote:
A combination of potential and actual energies Martin? Calculating the totality of impact that it may have? Or am I mistaken in my understanding of this?

Just to clarifiy on this post.

When designing a pumping system you will do all calculations in Head (Meters).

http://fachlerio.files.wordpress.com/2012/05/b3.jpg

This equation holds true for any fluid whether it is moving or not.

That constant is the same fluid at any other location with the same equation but with additions like losses from fricition or pump power added to the liquid.

If it is starting to click at this moment then you are correct. The main purpose of a pump in a closed loop is basically to keep this equation constant.

Hence... in a closed loop at any given point the velocity function is the same. The pressure function is changing and elevation function is changing. The pump added power is to equalize this number and overcome any losses.

The pump the only thing effecting the velocity function in the closed loop if all piping is the same width.

If you remember we stated that different pump designs give out different Head/flow even at the same RPM. This design factor changes the velocity function.

In very simple words. Get the flow you want and the head function will sort out it's own meaning.





Off topic... Bah I need to insert a flow meter in my loop now instead of just watching the height of the water!
Hydra
(13 items)
 
Hydra 2
(23 items)
 
 
CPUMotherboardGraphicsRAM
Intel i7 3930k @4.4ghz@1.44v Asrock X79 Fataility Champion 3xInno3d GTX 780 with Full Nickel EK Waterblocks Gskill RipjawsX 4x4GB@2400mhz 11-11-11-31 
Hard DriveCoolingMonitorKeyboard
2xSamsung 830 256gb Custom H2O - 3x XSPC RX360 with EK Blocks Surround Gaming 3x1440@96hz (X-Star) Cooler Master Trigger Red Caps 
PowerCaseMouseAudio
Seasonic 1250 watt Danger Den Double Wide 21 Roccat Kone+ Creative ZX (Incoming) 
Audio
Roccat Kave 5.1 Headphones 
CPUMotherboardGraphicsGraphics
4930k Asus Extreme Black Edition Sapphire R9 290x Sapphire R9 290x 
GraphicsGraphicsRAMHard Drive
Sapphire R9 290x Sapphire R9 290x Gskil TridentX 2400mhz Samsung 830 256GB 
Hard DriveCoolingCoolingCooling
Samsung 840 500GB EK Supramacy 4x EK Blocks 6x EK XTX480 
CoolingOSMonitorMonitor
EK XT360 Windows 8.1 BenQ 3201 32in 4k IPS Qnix 32in 1440 MVA 
MonitorKeyboardPowerCase
Qnix 32in 1440 MVA Coolermaster Storm Trigger 2x Seasonic 1250watt Case Labs TH10 
MouseAudioAudio
Roccat Kone Creative ZX AKG K7XX 
  hide details  
Reply
Hydra
(13 items)
 
Hydra 2
(23 items)
 
 
CPUMotherboardGraphicsRAM
Intel i7 3930k @4.4ghz@1.44v Asrock X79 Fataility Champion 3xInno3d GTX 780 with Full Nickel EK Waterblocks Gskill RipjawsX 4x4GB@2400mhz 11-11-11-31 
Hard DriveCoolingMonitorKeyboard
2xSamsung 830 256gb Custom H2O - 3x XSPC RX360 with EK Blocks Surround Gaming 3x1440@96hz (X-Star) Cooler Master Trigger Red Caps 
PowerCaseMouseAudio
Seasonic 1250 watt Danger Den Double Wide 21 Roccat Kone+ Creative ZX (Incoming) 
Audio
Roccat Kave 5.1 Headphones 
CPUMotherboardGraphicsGraphics
4930k Asus Extreme Black Edition Sapphire R9 290x Sapphire R9 290x 
GraphicsGraphicsRAMHard Drive
Sapphire R9 290x Sapphire R9 290x Gskil TridentX 2400mhz Samsung 830 256GB 
Hard DriveCoolingCoolingCooling
Samsung 840 500GB EK Supramacy 4x EK Blocks 6x EK XTX480 
CoolingOSMonitorMonitor
EK XT360 Windows 8.1 BenQ 3201 32in 4k IPS Qnix 32in 1440 MVA 
MonitorKeyboardPowerCase
Qnix 32in 1440 MVA Coolermaster Storm Trigger 2x Seasonic 1250watt Case Labs TH10 
MouseAudioAudio
Roccat Kone Creative ZX AKG K7XX 
  hide details  
Reply
post #22 of 74
Yes, The pump takes in electrical energy and converts it into the two forms of hydraulic energy. Flow rate is the kinetic piece and water pressure is potential side and some of that is the tubing and parts expanding under pressure. Pressure energy is then lost through friction along the way in all the blocks, etc (pressure) drop but the kinetic energy (flow rate) remains constant throught the loop.

These small pumps are very inefficient ant typicall net 15-25% efficiency which means a greater portion is lost through heat (heat dump), vibration, and noise. Pressure energy is also what is lost and converted to heat and noise at the blocks and such.

In a closed loop such as we have we can discard the static pressure energy because what goes up also comes back down. This leaves us with primarily flow and pressure energy, flow rate energy remains constant throughout a loop, but pressure and pressure energy is at it's maximum directly after the pump and back to zero at the pump inlet.

It is the pumps constant replenishing of pressure differential that ultimately does all the work to make the fluid move but both hebad and flow are critical in determining the right pump and why we see pump head vs flow curves. There is a careful tradeoff and relationship between the two specific to the pump.

I like to think of it at pressure gain (pump) and pressure drop (loop restriction), but pressure is a fundamental of how it works.

It doesn't have to be all that complex when selecting a pump either. Use the 1GPM rule of thumb and simply compare pressure gain vs loss at 1GPM. If you have more gained by the pump than lost by the parts, you will net greater than 1GPM.

Want to simplify it further, 6-10watts worth of pump is typically enough for lower restriction setup, CPU and GPU type loop. Want some extra flexibility, get a 18-25watt pump. Want every last bit? Get 50watts worth of pump or two smaller 18-25watt pumps in series. More will not necessarily yield better temps as you start seeing more heat dump loss than flow rate gain.

I have yet to build a loop that would not be satisfied by you typical DDC w/ to or D5, but I have run up to three for fun. Building extremes is a big part of the fun too..smile.gif I also enjoy working with the smaller 6-8 watt pumps as I know with good pressure drop information, you can actually design and operate a fairly complex loop with proper planning efforts.

It is all fun and interesting stuff...smile.gif
Edited by Martinm210 - 2/3/13 at 10:57pm
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
post #23 of 74
Myself I just love the DDC ones... Extremlly good head at low rpm. Got a MCP35x2 with its bay in my rig just for the giggles... ~15% PWM and it's serving the whole right nicely! biggrin.gif

Time to talk about Rads:

First rule of thumb.... Your rads are as good as the cold air you blow through them. Forget about the small things like RPM/Fins... First question that pops up.... cold air? you haz it? If yes then poke fun at the rest of the variables of picking a rad. I mainly focus on tests done all ready on rads and I highly thank martin for providing those charts for wattage numbers on rads.

Lets talk about Ambient temperature and its importance to normal heat transfer (not forced heat transfer or more termoinology correct forced conduction):
Ambient temperature dictates the lowest temperature you can go to.
You will never go below ambient temperature if you are using only fans and rads.
The bigger the difference between the two... the higher wattage the rads will output.
It's a long equation but I remember that a difference between two temperatures is and will be always the major factor in cooling.

Here is an examble:
Your ambient temperature is 20'C... your coolant is 30'C. A difference of 10'C.
Your ambient temperature is 20'C.... your coolant is 35'C. A difference of 15'C.

In the second condition your rads will output more heat from the system. Don't have the numbers but I writing this as fast as I can between breaks at work.

As the difference shrinks... the harder for the rads to output heat (I am talking about same fans and rads between the two conditions). Thus having 4'C difference between the ambient and coolant is considered extremlly good.

Now we talk about the conditions between the coolant and the blocks.... the cooler the coolant the better the cooling on the blocks..... Did it click yet?
Reduce your ambients => Cools better on the blocks.
Irrelevent of the rads used at the time.
You want a fast 10'C drop on parts... reduce your ambients by 10'C. Rads are a means of taking heat from the liquid to the atmosphere... if theirs not much of a difference between them then the rads are useless.

I actually did an experiment once... Had the fan running at 900 rpm from 1600 rpm... and the flow at about 0.7gpm but I reduced the ambient temperatures by 15 degrees (it was a 5'C night at the time). I had better figures then when I was 20'C Ambient.

When I go on these forums everybody is smacking that rad and this rad.... They forgot that at maximum variation they will have 20% difference..... while By just reducing the ambients by 5'C you gain a far better cooling effect.

Also rad selection is outrageous... everybody is thinking... MORE RADS MORE RADS.... slow down.... 240mm rad per OCed block is good enough... go 360 for Overkill but then you start dropping into dimenshing returns. "Semi-rant there".

Bottom line is... got enough rads? go for a better AC in that room... please?
Hydra
(13 items)
 
Hydra 2
(23 items)
 
 
CPUMotherboardGraphicsRAM
Intel i7 3930k @4.4ghz@1.44v Asrock X79 Fataility Champion 3xInno3d GTX 780 with Full Nickel EK Waterblocks Gskill RipjawsX 4x4GB@2400mhz 11-11-11-31 
Hard DriveCoolingMonitorKeyboard
2xSamsung 830 256gb Custom H2O - 3x XSPC RX360 with EK Blocks Surround Gaming 3x1440@96hz (X-Star) Cooler Master Trigger Red Caps 
PowerCaseMouseAudio
Seasonic 1250 watt Danger Den Double Wide 21 Roccat Kone+ Creative ZX (Incoming) 
Audio
Roccat Kave 5.1 Headphones 
CPUMotherboardGraphicsGraphics
4930k Asus Extreme Black Edition Sapphire R9 290x Sapphire R9 290x 
GraphicsGraphicsRAMHard Drive
Sapphire R9 290x Sapphire R9 290x Gskil TridentX 2400mhz Samsung 830 256GB 
Hard DriveCoolingCoolingCooling
Samsung 840 500GB EK Supramacy 4x EK Blocks 6x EK XTX480 
CoolingOSMonitorMonitor
EK XT360 Windows 8.1 BenQ 3201 32in 4k IPS Qnix 32in 1440 MVA 
MonitorKeyboardPowerCase
Qnix 32in 1440 MVA Coolermaster Storm Trigger 2x Seasonic 1250watt Case Labs TH10 
MouseAudioAudio
Roccat Kone Creative ZX AKG K7XX 
  hide details  
Reply
Hydra
(13 items)
 
Hydra 2
(23 items)
 
 
CPUMotherboardGraphicsRAM
Intel i7 3930k @4.4ghz@1.44v Asrock X79 Fataility Champion 3xInno3d GTX 780 with Full Nickel EK Waterblocks Gskill RipjawsX 4x4GB@2400mhz 11-11-11-31 
Hard DriveCoolingMonitorKeyboard
2xSamsung 830 256gb Custom H2O - 3x XSPC RX360 with EK Blocks Surround Gaming 3x1440@96hz (X-Star) Cooler Master Trigger Red Caps 
PowerCaseMouseAudio
Seasonic 1250 watt Danger Den Double Wide 21 Roccat Kone+ Creative ZX (Incoming) 
Audio
Roccat Kave 5.1 Headphones 
CPUMotherboardGraphicsGraphics
4930k Asus Extreme Black Edition Sapphire R9 290x Sapphire R9 290x 
GraphicsGraphicsRAMHard Drive
Sapphire R9 290x Sapphire R9 290x Gskil TridentX 2400mhz Samsung 830 256GB 
Hard DriveCoolingCoolingCooling
Samsung 840 500GB EK Supramacy 4x EK Blocks 6x EK XTX480 
CoolingOSMonitorMonitor
EK XT360 Windows 8.1 BenQ 3201 32in 4k IPS Qnix 32in 1440 MVA 
MonitorKeyboardPowerCase
Qnix 32in 1440 MVA Coolermaster Storm Trigger 2x Seasonic 1250watt Case Labs TH10 
MouseAudioAudio
Roccat Kone Creative ZX AKG K7XX 
  hide details  
Reply
post #24 of 74
Quote:
Originally Posted by 111ch1 View Post

A combination of potential and actual energies Martin?

No, not potential energy. Rather static energy. Just because there is no movement, does not mean there is no energy.
Quote:
Waiting for more information as I like where this thread is going. Quality Content. Indeed. Just want to say subscribed and researching as I wait for more replies. I really liked the clarity with which you made your post King4x4, should make that a guide for beginning water coolers(like myself) instead of this, 'buy these parts based on reviews then connect and voila!' I keep finding.
I plan on reading into this some more.

"I dare do all that may become a man.
Who dares more is none."
-Hamlet
Dark Vader
(6 items)
 
  
CoolingCoolingCoolingCooling
39 X AP-45 2150 rpm Gentle Typhoon fans Lamptron CW611 FrozenQ Dual Bay Reservoir 2x XT45 480. 1x UT60 480 ~ 1x UT60 360 ~ 1x UT... 
OSCase
Linux Ubuntu CaseLabs STH10 
  hide details  
Reply
Dark Vader
(6 items)
 
  
CoolingCoolingCoolingCooling
39 X AP-45 2150 rpm Gentle Typhoon fans Lamptron CW611 FrozenQ Dual Bay Reservoir 2x XT45 480. 1x UT60 480 ~ 1x UT60 360 ~ 1x UT... 
OSCase
Linux Ubuntu CaseLabs STH10 
  hide details  
Reply
post #25 of 74
Thread Starter 
You've misunderstood, and the people that wrote those articles have been sloppy with their language. Try reading some university slides from a decent university.

A fluid be definition, is a medium with no fixed molecular structure that will move to fill any vessel it placed it. It moves across pressure gradients from high pressure to low pressure.

All of these forms of energy you keep talking about as though the exist in the fluid as a consequence of its pressure, they do not. The exist in other locations in the system, in the case of an ACTUAL head of water, it is gravitational potential energy. In a vessel or hose, elastic potential in aforementioned.

Take Bernoulli's No.1 hit. 1/2*rho*v^2 + P + rho*g*z = constant.
We can express each of these elements, in terms of energy.
1/2*rho*v^2 is the kinetic energy OF the system.
rho*g*z is the potential energy OF the system.
and P is the work done ON the system. Not of, on.

A fluid cannot store energy. It doesn't have a rigid molecular structure in which to store energy, if I poke it, it doesn't bounce back, it just fills whatever space I put it in. Its pressure only exists as a function of its internal kinetic energy (or temperature if you prefer to think of that way).

EDIT Or a external force acting upon it such as weight from fluid above.
Edited by YowZ - 2/4/13 at 1:37am
 
Phantom Power
(15 photos)
 
CPUMotherboardGraphicsGraphics
Intel Core i7 860 ASUS Maximus III Gene NVidia GTX560Ti 2GB NVidia GTX560Ti 2GB 
RAMHard DriveHard DriveHard Drive
Corsair Vengence OCZ Vertex 3 WD Scorpio Blue 320GB X 2 RAID0 ARRAY OCZ Vertex 4  
Optical DriveCoolingCoolingCooling
OEM DVD-RW EK Supreme HF EK XTX 360 Phobya 18W Pump 
CoolingOSMonitorMonitor
EK 250 Advanced  Windows 8 Professional ASUS VG236H Acer T232HL 
KeyboardPowerCaseMouse
Filco Majestouch 2 Ninja OCZ ZS 750 NZXT Phantom White w/ Red Accent Custom Two Tone RAT 5 
Mouse PadAudio
Mionix Sargas 900 ASUS Xonar DGX 
  hide details  
Reply
 
Phantom Power
(15 photos)
 
CPUMotherboardGraphicsGraphics
Intel Core i7 860 ASUS Maximus III Gene NVidia GTX560Ti 2GB NVidia GTX560Ti 2GB 
RAMHard DriveHard DriveHard Drive
Corsair Vengence OCZ Vertex 3 WD Scorpio Blue 320GB X 2 RAID0 ARRAY OCZ Vertex 4  
Optical DriveCoolingCoolingCooling
OEM DVD-RW EK Supreme HF EK XTX 360 Phobya 18W Pump 
CoolingOSMonitorMonitor
EK 250 Advanced  Windows 8 Professional ASUS VG236H Acer T232HL 
KeyboardPowerCaseMouse
Filco Majestouch 2 Ninja OCZ ZS 750 NZXT Phantom White w/ Red Accent Custom Two Tone RAT 5 
Mouse PadAudio
Mionix Sargas 900 ASUS Xonar DGX 
  hide details  
Reply
post #26 of 74
Quote:
Originally Posted by YowZ View Post

You've misunderstood, and the people that wrote those articles have been sloppy with their language. Try reading some university slides from a decent university.


I am sorry I misunderstood I am learning (thank you for you assistance YowZ). Currently going to school but not taking any classes on this currently so I apologize for learning curve I am piecing this together from what I remember from basic classes therefore unenlightened/uneducated/worthless/failing in regards to this at this time (why I am thankful for all of this knowledge and help)

Quote:
Originally Posted by King4x4 View Post

When designing a pumping system you will do all calculations in Head (Meters).

http://fachlerio.files.wordpress.com/2012/05/b3.jpg

This equation holds true for any fluid whether it is moving or not.


Could you clarify what each of those variables are?

Quote:
Originally Posted by YowZ View Post

A fluid be definition, is a medium with no fixed molecular structure that will move to fill any vessel it placed it. It moves across pressure gradients from high pressure to low pressure....
A fluid cannot store energy. It doesn't have a rigid molecular structure in which to store energy, if I poke it, it doesn't bounce back, it just fills whatever space I put it in. Its pressure only exists as a function of its internal kinetic energy (or temperature if you prefer to think of that way).


I get this part, as the temperature changes its activity internally increases as molecules in liquid state move around creating kinetic energy and therefore moving to lower pressure.

Quote:
Originally Posted by King4x4 View Post

That constant is the same fluid at any other location with the same equation but with additions like losses from fricition or pump power added to the liquid.

If it is starting to click at this moment then you are correct. The main purpose of a pump in a closed loop is basically to keep this equation constant.
Quote:
Originally Posted by YowZ View Post

All of these forms of energy you keep talking about as though the exist in the fluid as a consequence of its pressure, they do not. The exist in other locations in the system, in the case of an ACTUAL head of water, it is gravitational potential energy.

Take Bernoulli's No.1 hit. 1/2*rho*v^2 + P + rho*g*z = constant.
We can express each of these elements, in terms of energy.
1/2*rho*v^2 is the kinetic energy OF the system.
rho*g*z is the potential energy OF the system.
and P is the work done ON the system. Not of, on.
.
Quote:
Originally Posted by Martinm210 View Post

Yes, The pump takes in electrical energy and converts it into the two forms of hydraulic energy. Flow rate is the kinetic piece and water pressure is potential side and some of that is the tubing and parts expanding under pressure. Pressure energy is then lost through friction along the way in all the blocks, etc (pressure) drop but the kinetic energy (flow rate) remains constant throught the loop...

In a closed loop such as we have we can discard the static pressure energy because what goes up also comes back down. This leaves us with primarily flow and pressure energy, flow rate energy remains constant throughout a loop, but pressure and pressure energy is at it's maximum directly after the pump and back to zero at the pump inlet.


What I take from this (since I do not know all the variables but looking up Bernoouli at this moment) is that you want a velocity(flow rate) for the liquid to have it move through creating its flow but in order to do so you need to account for the resistance or pressure changes in meters as well as static by creating head to propel it. Is this correct?

Everything to be accounted for then(te be all inclusive) is the pressure/resistance in radiator/blocks due to curvature and increased distance to travel, height changes in the system, length of tubing, resistance/pressure in the reservoir, and the friction of traveling through all the aforementioned items?

Quote:
Originally Posted by Martinm210 View Post

These small pumps are very inefficient ant typicall net 15-25% efficiency which means a greater portion is lost through heat (heat dump), vibration, and noise. Pressure energy is also what is lost and converted to heat and noise at the blocks and such.

It doesn't have to be all that complex when selecting a pump either. Use the 1GPM rule of thumb and simply compare pressure gain vs loss at 1GPM. If you have more gained by the pump than lost by the parts, you will net greater than 1GPM.

Want to simplify it further, 6-10watts worth of pump is typically enough for lower restriction setup, CPU and GPU type loop. Want some extra flexibility, get a 18-25watt pump. Want every last bit? Get 50watts worth of pump or two smaller 18-25watt pumps in series. More will not necessarily yield better temps as you start seeing more heat dump loss than flow rate gain.


Let's say I wanted to negate that energy loss by going with a better pump, are there better units available in some other commercial field? In addition to that, lets say that a 20 watt pump is required, should a 25 pump suffice as long as there is a 1GPM flow running through the system so that there is a constant diffusion of temperature through thermal conductivity? Is there a balance to minimize heat dump? (i.e. running the pump at X% of capability?)

Side note here also... I currently planned on having the pump at the base of the system for ease of filling with coolant but should I be worried of heat generated by this on my setup?

Wall of text on proposed rig I have in mind (Trying to water cool to maximize performance trade off with cost. I know TEC / Phase change cost a lot to operate while air conditioning also costs a lot to operate so sort of out of the question for radiators but believe I can get steady 5Ghz clock with something like this.)
post #27 of 74
A single DDC or D5 variant will work fine. DDCs are more electrically efficent and have a slightly stronger curve when coupled with an aftermarket top, but the D5s are more silent and basically watercooled so no heat buildup issues. I like cooling all DDCs myself with at least good air flow blowing under the base. Some good temp probes to measure air in and water temps will serve you well in seeing how things are doing. Fresh cold ambient intake air is the more critical part as King noted, but I do also like to go overkill on the rad as long as I can manage good cold intake.

You can also do some rudimentary flow rate tests and see where you fall. Dual pumps in series usually only relates to a 30% or so increase in flow rate max, so using some data like Strens roundup at extremerigs can give you a useful look at what flow rate gains remain.

My favorite setups are with PWM controllable pumps using speedfan and a curve to throttle both fans and pump. I also allow throttling of the CPU.

3930K drops down to about a 25w heat source at idle, and up to about 180w at full load. A 2x120 rad with 1800 rpm capable fans is plenty capable of coling that with good case airflow, so one 120mm rad section per 90-100w is not a bad rad rule of thumb for those fan speeds, but more doesn't hurt.

I like shooting for a 5C water to air delta, but that is pretty hard to do inside a case. 10C is more typical and even 15C can still get you by.

There is no absolute minimum flow rate either, each block is a little differnet in how well it handles low flow, but we know it is possible to rund down in the .5 GPM areas with still acceptable performace. 1 GPM is more of a rule to help ensure good bleeding and plenty for performance. Inverted double core rads and large ID tubing can have bleeding problems if flow rates get too low.

Unfortunately a majority of products out there do not include pressure drop curves info, so no book or theory is going to help us design or calculate a flow. It may be interesting, but without the pressure drop data the best we can do is try it, measure some actual flow rates, and add some temperature probes to see how it is performing.

Hope that helps..
Martin
Edited by Martinm210 - 2/4/13 at 6:02am
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
post #28 of 74
Quote:
Originally Posted by YowZ View Post

You've misunderstood, and the people that wrote those articles have been sloppy with their language. Try reading some university slides from a decent university.

A fluid be definition, is a medium with no fixed molecular structure that will move to fill any vessel it placed it. It moves across pressure gradients from high pressure to low pressure.

All of these forms of energy you keep talking about as though the exist in the fluid as a consequence of its pressure, they do not. The exist in other locations in the system, in the case of an ACTUAL head of water, it is gravitational potential energy. In a vessel or hose, elastic potential in aforementioned.

Take Bernoulli's No.1 hit. 1/2*rho*v^2 + P + rho*g*z = constant.
We can express each of these elements, in terms of energy.
1/2*rho*v^2 is the kinetic energy OF the system.
rho*g*z is the potential energy OF the system.
and P is the work done ON the system. Not of, on.

A fluid cannot store energy. It doesn't have a rigid molecular structure in which to store energy, if I poke it, it doesn't bounce back, it just fills whatever space I put it in. Its pressure only exists as a function of its internal kinetic energy (or temperature if you prefer to think of that way).

EDIT Or a external force acting upon it such as weight from fluid above.

I understand the "nearly"incompressible fluid part, and not arguing that the tubing in our loops probably store much of this energy, but the energy is there and produced by the pump. A block can not produce energy, it can only consume energy through friction. The pump is the only item creating hydraulic energy in a loop.

Understanding pressure head and head loss is the basis for designing any pump or hydraulic system so saying a pump doesn't create pressure will only cause more confusion than help.. Friction is an energy loss and head loss is the tool we use to measure friction and pump head is the tool we used to understand how well a pump can overcome that friction. Without Head, we are lost.

I appeciate the technical discussion and information very much.
Cheers!
Martin
Edited by Martinm210 - 2/4/13 at 8:06am
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
post #29 of 74
Quote:
Originally Posted by 111ch1 View Post

Everything to be accounted for then(te be all inclusive) is the pressure/resistance in radiator/blocks due to curvature and increased distance to travel, height changes in the system, length of tubing, resistance/pressure in the reservoir, and the friction of traveling through all the aforementioned items?
Let's say I wanted to negate that energy loss by going with a better pump, are there better units available in some other commercial field? In addition to that, lets say that a 20 watt pump is required, should a 25 pump suffice as long as there is a 1GPM flow running through the system so that there is a constant diffusion of temperature through thermal conductivity? Is there a balance to minimize heat dump? (i.e. running the pump at X% of capability?)
If you are worried about heat dump from the pump, get a DCC as they dump their heat in air while D5 pumps dump their heat in the water. That's why a DCC pump feels like it's running hotter.

There is pros and cons to both of these characteristics.
Dark Vader
(6 items)
 
  
CoolingCoolingCoolingCooling
39 X AP-45 2150 rpm Gentle Typhoon fans Lamptron CW611 FrozenQ Dual Bay Reservoir 2x XT45 480. 1x UT60 480 ~ 1x UT60 360 ~ 1x UT... 
OSCase
Linux Ubuntu CaseLabs STH10 
  hide details  
Reply
Dark Vader
(6 items)
 
  
CoolingCoolingCoolingCooling
39 X AP-45 2150 rpm Gentle Typhoon fans Lamptron CW611 FrozenQ Dual Bay Reservoir 2x XT45 480. 1x UT60 480 ~ 1x UT60 360 ~ 1x UT... 
OSCase
Linux Ubuntu CaseLabs STH10 
  hide details  
Reply
post #30 of 74
With Regards to King4x4's earlier comments about radiators I would just like to check my understanding is correct. For a loop there are two main types of heat interface the CPU(or other component)/Coolant interface and the Coolant/Air interface and at each of these there will be a delta T. Now for a given loop these temperature differences will remain constant. Lowering the ambient air temp causes the coolant temp to decrease as the Air/Coolant differential stays constant and then the CPU (or component) will also decrease in temp as the CPU/Coolant delta T must also remain constant. Is this correct?
Now when more radiator area is added to the same loop the Air/Coolant delta T is reduced, lowering the water temps (as the ambient Air temp stays the same) and subsequently lowering the temp of the CPU and the Coolant/CPU delta T is constant, it is also my understanding that adding more radiators has diminishing returns as the size of the delta T decreases while in theory lowering the ambient has linear returns. I hope someone can correct me if this is wrong, but if this is correct is it possibly more economical for people with large air/coolant delta T's to add more radiator space before lowering the ambient air temps?
I feel I should add I have no practical experience of water cooling just some theoretical knowledge of thermal dynamics from my degree, which is surprise, surprise…. Engineering (Civil).
The Kestrel
(9 items)
 
  
CPUMotherboardGraphicsRAM
i5 3570k Gigabyte Z77X-UD3H Asus 7870 Direct CUII Samsung Green 
Hard DriveCoolingOSPower
Kingston Hyper x 120Gb SSD Xigmatek Dark Knight II Widows 7 64bit Seasonic M12II 650W 80+ Modular PSU 
Case
NZXT Phantom 
  hide details  
Reply
The Kestrel
(9 items)
 
  
CPUMotherboardGraphicsRAM
i5 3570k Gigabyte Z77X-UD3H Asus 7870 Direct CUII Samsung Green 
Hard DriveCoolingOSPower
Kingston Hyper x 120Gb SSD Xigmatek Dark Knight II Widows 7 64bit Seasonic M12II 650W 80+ Modular PSU 
Case
NZXT Phantom 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › A few words about pumps and flow rate.