Is it possible to have too much flow?

[DoGMaN]

As the title says. Is it possible to have too much flow in a watercooling loop? I just swapped out my Liang D4 for an Ehiem 1260. Specs are (off the top of my head) 583 gph ~12ft head. Only thing hooked up at the moment is my rad (an old econoline van heater core 5/8" in/out) and an EK Supreme water block.

[Eddie3dfx]

It is possible to have too much flow (as it everything just can't handle the pressure and pops)
You are not even close to being there yet.

[nikhsub1]

As the title says. Is it possible to have too much flow in a watercooling loop? I just swapped out my Liang D4 for an Ehiem 1260. Specs are (off the top of my head) 583 gph ~12ft head. Only thing hooked up at the moment is my rad (an old econoline van heater core 5/8" in/out) and an EK Supreme water block.
I hate to be the one to tell you this but the 1260 is an awful pump for water cooling... it draws (IIRC) 60W of power, of which ~50W will be dumped into the loop. Not only that, your old D4 has more head and way less heat dump. You just went backwards in performance :down:

As far as too much flow, theoretically no there can never be too much. Realistically, yes there is a point to where you start to degrade performance with added heat to the system to achieve this flow. You have managed to do this just by choosing the wrong pump.

[Fr3ak]

nik said it all. At some point, temperatures won't get better, because the cooler can't dissipate more heat to the water. That point depends on the individual block.

[eXa]

If u look at the flow by itself then u cant really get too much. But to get more flow u need a bigger pump, and a bigger pump will add more heat to the loop. So at a point the performance will get worse since the little added flow aint enough to outperform the added heat from the pump.

[Martinm210]

Yeah heat dump is the factor, but I suppose if you wanted to think outside of the pumping variable there's also frictional heat. Pressure drop is simply a measurement of energy, and where and what form of energy does that lost pumping water horsepower go when it's lost through restriciton in the block. It is lost in the water, so not only do you have heat dump in the pump, you have a miniscule amount of heat being added to the water throughout the loop.

So even though the pump is really the one that matters, there would also be some sort of diminishing return if pumps didn't matter where heat dumped by the water's frictional head loss began to overcome the thermal efficiency gained by more flow rate across the block. There is always a limit...:D

[Shocker003]

Well i don´t think there is anything like too much flow. They have pointed out all the faults of a powerfully pump, but i still believe the heat dumped by a pump(Alphacool 1510 running @24v) can be countered by the number of less restrictive rad in a loop(2X MCR320). At the end it will still be a waste of money high electric bills. I had an alphacool AP 600 12v, it was a weak pump with 1.6m head, i ain´t so sure. I was adviced to upgrade to a better pump(Alphacool 1510) for better temp and i ended up with worse temps. Well i have learnt my lesson, it runs at 12v and i still added a BI 120 X-Flow to help out with heat dump.

[nikhsub1]

Well i don´t think there is anything like too much flow. They have pointed all the faults of a powerfully pump out, but i still believe the heat dumped by a pump(Alphacool 1510 running @) can be countered by an additional less restrictive rad(2X MCR320).
This reminds me of how much I miss the old days, sigh.

http://www.xtremesystems.org/forums/showthread.php?t=136678

[skinnee]

This reminds me of how much I miss the old days, sigh.

http://www.xtremesystems.org/forums/showthread.php?t=136678

I lurked that thread for the duration, go to read again though!

[DoGMaN]

I hate to be the one to tell you this but the 1260 is an awful pump for water cooling... it draws (IIRC) 60W of power, of which ~50W will be dumped into the loop. Not only that, your old D4 has more head and way less heat dump. You just went backwards in performance :down:

As far as too much flow, theoretically no there can never be too much. Realistically, yes there is a point to where you start to degrade performance with added heat to the system to achieve this flow. You have managed to do this just by choosing the wrong pump.

Yah I believe it is 60w. It is only a temporary thing, but I didn't want to cause any issues by changing to what I perceived to be a "more powerful" pump. I do understand what you are saying though :D. My D4 is out getting milled so it should be back in 3-4 days.

[Martinm210]

I agree shocker, the miniscule amount amount of heat a DDC3.2 puts out can easily be countered by an additional rad...

You'll have just that, so would you mind testing it with one pump on and then the other turned on?

That would be nice to have another person try it.:up:

Here was Andy's testing with one PA 120.3

Here an example of what happens when you use 1 pump/2 pumps (Liang D-5's)with a EK Supreme. I;ve also ran a test over 1.5 hours where I cycled the 2nd pump on and off in 20 minute intervals, but I haven't charted it yet. It will take me a while.



Colums are as follows, sorry having trouble formating excel to plain test.
test version/Avg-Air in/Avg-Flu In/ Air Fluid Delta-T/Air.Core Delta T/Core Temp/Adj. Core Temp (to air temp from first test)

EK v1| 24.7594| 29.1987| 4.3964| 33.2388| 57.9982| 57.9982 2 pump
EK v2| 24.6843| 29.1935| 4.5053| 33.1889| 57.8732| 57.9483 2 pump
EK v3| 25.2892| 29.2135| 3.8970| 32.9408| 58.2300| 57.7002 1 pump

It's repeatable also, but I still need to run several more 1 pump trends. Sometimes you can see a +/- 0.2c swing, just depends.

andyc

He is showing with the EK Supreme and a PA120.3, that he lost .2C using dual D5s. That's pretty consistent with my quick and dirty testing I did the other day on a different block using the TFC480 and YLD12SM12's with shrouds.

I suspect you might chase it out, but it would take two or three triple rads to offset the difference in heat(depends on the fans used) and you're probably talking tenths of a degree if that.

[Snyxxx]

I have just finished my home test of a single DDC-2 (old school orange impeller) versus two of them in series. I did a bucket flow test and got just over 1.0 gpm with one pump and just over 1.5 gpm with both in series. This matched pretty well with Martin's flow spreadsheet. I am basically using my signature stuff with another PA 120.1 ADDED.

I was using my bigNG to monitor ambient and water temperatures as well as Everest 4.5. I did P95 small FFTs and then measured idle after sufficient time.

Results surprised me. No measurable difference. We are talking .5 to 1°C between ALL values such as GPU1, GPU2, SPP, MCP and the cores.

I will still keep them in series for redundancy as one of them had troubles starting from a cold boot.

[eXa]

This reminds me of how much I miss the old days, sigh.

http://www.xtremesystems.org/forums/showthread.php?t=136678

Idd...

[fart_plume]

Yes, too high a flow rate would be detrimental, thats why they put thermostats in autos in not only controls the temps at which the coolant begins to flow, the hole in the thermostat also serves as an orifice to regulate flow. Too much flow not enough heat is absorbed by the coolant leading to over heat. Now where that maximum is for a computer I don't know.

[fart_plume]

you apparently know nothing about atoumotive cooling design. It can be appiled to any cooling system. If your flow rate is too high the coolant cannot absorb the heat, too low a flow rate problem is that your not removing the heat fast enough, resulting in the same end overheating, it has to be tuned like any other system for maximum effiency. If you don't believe me take your thermostat out of your car during the hottest part of the summer and watch the temps on the engine.

[Serpentarius]

As the title says. Is it possible to have too much flow in a watercooling loop? I just swapped out my Liang D4 for an Ehiem 1260. Specs are (off the top of my head) 583 gph ~12ft head. Only thing hooked up at the moment is my rad (an old econoline van heater core 5/8" in/out) and an EK Supreme water block.

while this guy mentioned TOO MUCH FLOW, how i wish my rig has TOO MUCH FLOW :ROTF:

actually, everyone welcomes too much flow ... but everyone also worries the process of getting "too much flow", which normally linked to too much heat dump

it's ok to have too much flow, for it helps the water turbulance which in turn pickups up heat faster ...

but most high power pump will also injects its heat dump which makes matters worst .. it's hard to balance between two

[fart_plume]

you better check again that hole that the button on a thermostat is sized for flowrate maximum when the engine is at maximum operating temperture, remove that button(the control vavle in the center of the thermostat) put it back in and it won't overheat but remove the thermostat completely and it will overheat. don't believe me try it yourself.

and a quote from wikipedia or a haynes manual is not a good reference for actual engineering. where your quote is essensially correct it does not say everything, it is a general reference quote for the layman, not for someone designing cooling systems.

[aspire.comptech]

you better check again that hole that the button on a thermostat is sized for flowrate maximum when the engine is at maximum operating temperture, remove that button(the control vavle in the center of the thermostat) put it back in and it won't overheat but remove the thermostat completely and it will overheat. don't believe me try it yourself.

and a quote from wikipedia or a haynes manual is not a good reference for actual engineering. where your quote is essensially correct it does not say everything, it is a general reference quote for the layman, not for someone designing cooling systems.

Right...

[fart_plume]

basically what I'm saying is you need to tune your cooling system for max effiency and performance, don't just rely on max flow rate, pressure will have an effect, block design, radiator design and even materials that each part is made of. you may find a high pressure and lower flow may improve or decrease the cooling systems performance, it must be tuned to work together.

[fart_plume]

LMAO...:rofl: I've been working on cars for years. You've got all the parts right, but not how they work or used to cool and engine. It's one of the most simple parts of an automobile engine. It closes up to restrict flow to the rad when the engine is cold, and opens to allow flow to the rad when the engine reaches operating temp or above. Once wide open, it's the fans job on the rad to keep temps under control. It's been that way for over 75 years.

If you're saying you know more than the Haynes manual, more power to you and your ride. Your original post, and the 2nd one basically said a thermostat restricts flow to keep the engine cool as some argument why to much flow is a bad thing for a WC'ing system. That's absurd and you basically don't under stand how a thermostat works in a cars engine.

It's not the hign or increased flow that's bad for a PC's water cooling system, it's the heat dump from the pumps providing that increased flow over an beyond what the rad can remove.

andyc

I and my family have been working on cars since the model T in fact grandfather was THE FIRST model T mechanic in Missouri.

and if you don't think that has an effect on cooling pull the thermostat and see for yourself. Just because they didn't teach you that a hole of a specific size can regulate flow rates doesn't mean it isn't so. Think about the orifices in a carb what do they do? control flow.

And I'll give a little more info on my background 30+ years in manufacturing and design, including water control systems.

[fart_plume]

let me put it this way then. you have a V8 engine typical setup approxamatly 2" hose from the rad to the water pump, and a 2" hose from the rad to the thermostat housing, and all these wide open passages through out the waterjacket if max flow rate was desirable in this system why would they design a thermostat with a 5/8" hole in it when ever they are quite capable of building one with a 2" hole to match the rest of the system?

[nikhsub1]

I'll not get involved with this any further... how did BillA put it? Feeding lard to the geese on a string?

[fart_plume]

Think outside of the box people, it's time to take WC'ing to the next level.

[Shocker003]

Yes, too high a flow rate would be detrimental, thats why they put thermostats in autos in not only controls the temps at which the coolant begins to flow, the hole in the thermostat also serves as an orifice to regulate flow. Too much flow not enough heat is absorbed by the coolant leading to over heat. Now where that maximum is for a computer I don't know.

In Africa we use to remove thermostats from auto cooling system to reduce the restriction it induces. Due to high climatic temps, the fans on these cars rad are equally made to be on all the time, unlike most cars that their fans kick in automatically. As for colder climatic regions, it helps to reduce flowrate to keep the engine in the best operatable temp but equally opens up it´s ventiles automatically to increase the flowrate when the engine coolant gets too warm. But as in pcs watercooling, high flowrate will not hurt your temp except when your pump is dumping extra heat into the loop. The best way to counter this extra heat dump is by under volting your pump or adding an extra rad to the said loop.

NB: During winter leave you car outside for over 10hrs with the engine turned off. Then start the car and try to acclerate from 0 to 150km. It will take your cold engine more time to reach that speed but when it warmed up stop and recheck the time it will take it to get to the same speed.;)

[twwen2]

Think outside of the box people, it's time to take WC'ing to the next level.

Good luck at the next level! :rolleyes:

[fart_plume]

You know I've seen a sub ambient temperature system ,that uses no chiller, why has no one here built one?

[fart_plume]

That's the attitude one expects from flat earthers...............

[aspire.comptech]

You know I've seen a sub ambient temperature system ,that uses no chiller, why has no one here built one?

What, a bong cooler?

Thats old news.

[majestik]

You know I've seen a sub ambient temperature system ,that uses no chiller, why has no one here built one?
TEC? Probably because they are :banana::banana::banana::banana:.

[fart_plume]

not a TEC system, a cooling tower system.

[majestik]

http://s95367906.onlinehome.us/XS/tower_cooling.jpg
Who wants one? I know I do!

[largon]

If your flow rate is too high the coolant cannot absorb the heat, Perhaps this is kinda like the same phenomenon as during winter (or any given season) the faster the wind the warmer the air feels?
Uh... Wait? Or was it the other ways around?

*poke*

[eXa]

@ Fart_plume

U have missunderstood the purpose of the thermostat. Unlike the cpus the engine likes to operate at a certain temperatur. Below(or above) this temperature u will wear the engine more. The thermostat closes so the little water around the cylinders can heat up faster and thus reach operating temp faster.

[twwen2]

:rofl: @ Majestik!

And you got the best avatar ever! I was gonna change to that pic tonight, you beat me to it! :p:

[C'DaleRider]

fart....I understand your family has been playing with cars since the Model T....you'd probably find lots of people's families around here have similar experiences. Unfortunately, you need to move your thinking into the modern era and quit thinking of 1950's cooling tech.

A common misconception is that if coolant flows too quickly through the system, that it will not have time to cool properly. However the cooling system is a closed loop, so if you are keeping the coolant in the radiator longer to allow it to cool, you are also allowing it to stay in the engine longer, which increases coolant temperatures. Coolant in the engine will actually boil away from critical heat areas within the cooling system if not forced through the cooling system at a sufficiently high velocity. This situation is a common cause of so-called "hot spots", which can lead to failures.

Years ago, cars used low pressure radiator caps with upright-style radiators. At high RPM, the water pump pressure would overcome the radiator cap's rating and force coolant out, resulting in an overheated engine. Many enthusiasts mistakenly believed that these situations were caused because the coolant was flowing through the radiator so quickly, that it did not have time to cool. Using restrictors or slowing water pump speed prevented the coolant from being forced out, and allowed the engine to run cooler. However, cars built in the past thirty years have used cross flow radiators that position the radiator cap on the low pressure (suction) side of the system. This type of system does not subject the radiator cap to pressure from the water pump, so it benefits from maximizing coolant flow, not restricting it.


As coolant flows through the system it absorbs heat from the engine parts that it comes in contact with. As it does this some of the coolant will boil and form tiny steam bubbles (absorbing a lot of heat in the process) on the internal engine surfaces. When these bubbles get larger they become a flow restriction and the flowing fluid pushes them away from the surface and that process starts over again.

The process is called the Nucleate Cooling Phase. When the coolant boiling point is too low or the flow rate is too slow, these bubbles can become too large and form steam pockets that insulate that surface from being cooled. This usually happens around the combustion chambers, the hottest parts of the engine. Once the steam pocket forms the surface will rise in temperature (even though the coolant is not overheating) and cause that part to overheat, which can cause detonation and / or other problems.


Different coolants require different minimum flow rates, but contrary to popular belief, you cannot make the coolant flow too fast. This rumor was started because people removed the thermostat to gain flow, because they had an over heating problem, and it only aggravated the problem. The real reason they ran into problems is that removing the thermostat also removes the restriction that builds pressure in the engine, so they gained flow, but reduced the boiling point of the coolant in the block.

Running a higher flow thermostat and a higher volume pump to maintain pressure, will give no such problems. If you think about it, making the coolant flow twice as fast will also make it flow though the engine twice as often, so there will be more even temperature across the engine. There has been, and still is, the rumor that of the coolant flows too fast, it will not have time to pick up heat. That is nonsense, as long as there is coolant contact a surface, the rate of heat transfer will be the same. Coolant that flows twice as fast also flows through the block twice as often.

The thermostat is simply there to get the engine's temperature up to optimal operating temperatures as quickly as possible, which in an internal combustion engine in a car is around 200F. Below that and the computer controlled engine operates in a closed loop setting, running the engine rich. If you remember the 1950's tech for this same system, it was a choke on the carburetor....it'd close off the air flow into the carb initially and slowly open as the engine warmed, essentially running the engine rich and leaning it out as the engine warmed as the choke opened up. This was a very crude version of what today's computer controls do by providing more fuel per injector pulse when the engine's cold and leaning out as the engine warms.

The thermostat's main job is to allow the engine to heat up quickly, and then to keep the engine at a constant temperature. It does this by regulating the amount of water that goes through the radiator. At low temperatures, the outlet to the radiator is completely blocked -- all of the coolant is recirculated back through the engine.

Once the temperature of the coolant rises to between 180 and 195 F (82 - 91 C), the thermostat starts to open, allowing fluid to flow through the radiator. By the time the coolant reaches 200 to 218 F (93 - 103 C), the thermostat is open all the way.

[fart_plume]

you still didn't answer the question, if it wasall about highest flow rate possible then why don't they make a thermostat with a 2" hole to match the rest of the system?

[Waterlogged]

you still didn't answer the question, if it wasall about highest flow rate possible then why don't they make a thermostat with a 2" hole to match the rest of the system?

Don't have time for a detailed answer right now but, flow rate inside a automotive cooling system is irrelevant.

[C'DaleRider]

Oh, and about the thermostat restricting flow.....in a sense it does indeed do that, farty. But the restriction it provides is not to drop flow at all but instead is to provide a way for the water pump to build pressure. A pump running without any restriction builds very little to no pressure......so the thermostat does indeed provide a way for the pump to build pressure inside the engine. You could retort that the radiator is another way for restriction to be added into the cooling loop, but modern crossflow radiators are extremely low restriction devices and provide little restriction to the coolant, so the thermostat is indeed needed for pressure buildup.....but you can indeed run successfully without a thermostat.

The problem is the coolant at that point. Running an engine sans thermostat and with water only could lead to overheating, as you alluded to....but modern coolants do not depend upon pure water. Instead some sort of glycol is added, thereby raising the boiling point of the coolant.

A water / ethylene glycol mixture will boil at a higher temp and resist steam pockets better than plain water, the down fall is that it has to have a higher flow rate, but that is easy to accomplish.

Another common form of coolant is propylene glycol, which has the highest boiling point and can run higher than 250° F (average temperature as seen on a gauge) without forming steam pockets, but it must flow at more than twice the speed of a water / ethylene glycol mixture (which means major changes to most cooling systems).

So, even without a thermostat in place, a water/glycol mix will have a boiling point higher than pure water. Add the restriction of a thermostat into the mix, which allows the pump to build higher pressures, and you can move the boiling point to upwards of close to 300F.

[fart_plume]

The automotive crap is a moot point , what I'm trying to do is to get you guys to think out of the box. What are things you haven't tried, how can you make the cooling systems better? Go beyond what the blocks are made of, beyond highest flowrates. would higer pressures in the loop improve cooling, coupled with a tuned flowrate for maximum heat transfer, I'm not saying slow the flow to a crawl or increasing to the point of no return, but a challenge to you to find ways to improve the performance of the systems. Be like a racing team, and not like a weekend hotrodder. and figure out how to tune a system for performance and not by simply undervolting or overvolting pumps, buying off the shelf parts. Get into the meat of the system and figure out how to make it work better, beyond it's off the shelf specs.

[NaeKuh]

*pulling fart's finger releaseing plume*

Everyone EVAC NOW!


http://s95367906.onlinehome.us/XS/tower_cooling.jpg
Who wants one? I know I do!

Pish... that tower aint tall enough for my monster rig.

:rofl:

[hotdun]

The automotive crap is a moot point , what I'm trying to do is to get you guys to think out of the box.

It's a moot point now????

Why don't you think outside the box and then relay that information to us "flat earthers".....if we can even comprehend it.....

[C'DaleRider]

Guess when each and every auto question farty posed was answered, the auto part became moot......very interesting. Evasive, but interesting.

[fart_plume]

Same attitude again, there are how many of you, don't you have any ideas to improve systems, that you haven't tried? Or are you simply afraid to air them for fear of rebuke by diehards to only this and this works? Over the last year watching this section nothing new has really been posted. nothing new tried. Just a rehash of don't mix metals tubing size, which pump is the best, same old stuff. Get those ideas out in the open that you want to try lets find some new ground.

[fart_plume]

the thing is I do understand the principles and hydrodynamics very well, and as with any system there IS a point of dimishing returns. The small pumpos used in computers may not be enough to reach that point ,yet. This section is in such a sad state of bling that the XTREME has gone comepletely out it. Go look at the phase change section, where the guys are tried to be on the bleeding edge. their concern is for performance not looks.

You wail about mixed metal systems and don't give any constructive tips on how to help a guy to prevent corrosion in his exsisting system. Who cares what it looks like as long as it works at peak performance? And where are the threads about trying to push the limits? There haven't been any for quite some time.

[Kurz]

There is no need for higher pressures in a Water loop.
When does water see higher boiling points?

Nor is higher flow really needed.
The temperatures we work at is actually pretty weird for the rest of the world.

We are cooling a part that doesn't get that hot in the first place.
Trying to take it down to ambient temperatures. 70C *Gpu's* to 25C *Ambient*

The only thing that really can benefit the loop is:
*Increasing the surface area of Radiators.
*Increasing heat conductivity of water.
*Increasing heat conductivity of the water blocks.

Everything else is pretty much diminishing returns.
Anything above 2 Gpm is moot.
Tubing is where they are supposed to be.

[NaeKuh]

if things were only that simple. :\

Then we all be using TT and innovotek.

So Kurz no. Theres a lot more your missing out on, and your only covering the middle base.

You forgot the icing, and the layers.

Also the price/performance on Air vs Water is already diminising returns, so you cant really say that to begin with.

[fart_plume]

Now we're getting somewhere!

Has anyone looked at biological systems(as in animals) to see the way they cool themselves to get ideas? I have and I do have an idea and a design.

[Kurz]

if things were only that simple. :\

Then we all be using TT and innovotek.

So Kurz no. Theres a lot more your missing out on, and your only covering the middle base.

You forgot the icing, and the layers.

Also the price/performance on Air vs Water is already diminising returns, so you cant really say that to begin with.

What am I missing?

Plus I was comparing Water to Water not Water to air.
If I wanted to be cheap I would've gone air.

[fart_plume]

This is what i'm talking about, water cooling should be better than air, but right now air can acheive temps similar to water. When that happens all wc'ing is,is bling. There is performance work to be done to get back on top.

[hotdun]

This is what i'm talking about, water cooling should be better than air, but right now air can acheive temps similar to water. When that happens all wc'ing is,is bling. There is performance work to be done to get back on top.

Not when you bring OC'ing into the mix. Then it's a different story....not bling....

[Eternalightwith]

Land dwellers cool off by evaporative cooling. In computer land it's called bong cooling. If done right you can get a degree or 3 below ambient.


Now we're getting somewhere!

Has anyone looked at biological systems(as in animals) to see the way they cool themselves to get ideas? I have and I do have an idea and a design.

[aspire.comptech]

For a non overclocked system at idle, the difference in temps is minimal but the delta between air and water at load on an overclocked system is where it really shines.

Air cooling is good for 99.9% of all computer users and they will get perfect performance from it, But when you start dumping more heat into the mix, they just can't keep up.

fart, all you keep doing is spout on and on about thinking outside the box, but you don't bring anything new or interesting to the table.


Land dwellers cool off by evaporative cooling. In computer land it's called bong cooling. If done right you can get a degree or 3 below ambient.

Your about 12 hours late...

http://www.xtremesystems.org/forums/showpost.php?p=3207363&postcount=36

[NaeKuh]

The only thing that really can benefit the loop is:
*Increasing the surface area of Radiators.
*Increasing heat conductivity of water.
*Increasing heat conductivity of the water blocks.

Surface area of radiators. yes, but what about flow drop? And CFM of fans?

Increasing heat conductivity of water: Well thats not possible at all, you can only make it worse, not make it better. Unless im not knowing about something.

Increase heat conductivity of the water blocks: Yes but how you going to do this? You know increasing surface area on the block is pointless, otherwise nikhsub1's stepping block test would of proved wrong otherwise. Also what about pressure? Pressure is not uniform, and some blocks benifit from the added pressure.

See your missing a TON of stuff. Its not as simple as you make it sound.

[fart_plume]

there more to biologolocal systems than simple evaporation, such as in homosapiens those fine hairs on your arms are part of that cooling system, get your thinking yet? No? then try this why are desert animals small and artic animals large?

[Xilikon]

there more to biologolocal systems than simple evaporation, such as in homosapiens those fine hairs on your arms are part of that cooling system, get your thinking yet? No? then try this why are desert animals small and artic animals large?

The hair on the arms (and other parts of the body) is there to regulate the temp and keep yourself warm. It's not there to help with cooling, sweating is doing the cooling job. the reason the body need to regulate the temp is because are warm blood organisms. Snakes is cold blood so it rely on the environment to regulate the body temp. This is the same principle as the car thermostat to regulate the engine temperature that you don't seem to grasp yet.

Regulation of temps != cooling performance.

/me get out of the rat nest...

[fart_plume]

try jack rabbits, mammals not reptiles, and those hairs do help with cooling by holding the sweat next to the skin so that when air moves through the hair you get more cooling effect from evaporation, than if the sweat simple rolls off.

[NaeKuh]

hey FP, i hate to throw a wrench at ya.

But you cant get evap occuring inside a closed loop, unless its due to cavitation. Which then its actually vapor stuck in your inlet.
:T

Your idea is good, if the waterblock was open and you could evap as it traveled, but in a closed loop, there is no such thing as evap, unless your talking about a bong cooler which requires an exhaust.

(sub note: the tubes are actually pourous, so you do lose coolant over time due to evap, and osmosis but not on a degree where it would cool the liquid)

Your idea of increasing surface area is good, however thats easily accomplished via other methods like glass beading, and cross slits or drilling.

You cant get evap occuring unless your on HWLabs RxN which never got launched.
Im guessing due to coolant loss and humidity.

[fart_plume]

ok, the thing is relate the two, items surface area vs mass percentage and hairs instead of fins or pins. to make a fin or pin you can only go so small before you can't go any smallerhowever you can draw a copper wire hair much smaller. due to it's small diameter you have a greater surface area percentage plus you can pack more hairs in a block than pins, giving more total surface area in the same space, there by increasing the heat removal potential for a cooling block.

[NaeKuh]

but your hair strands would need to be somewhat long then since you want surface area. And that would make the hairs also very brittle. The smaller it becomes, the greater chances of it breaking off, and lodging into your pump impeller :down:

Expecially when you have swiftech mystery jelly in your loop when you use Hydrex.
Fesser coolant would probably disolve those small fibers. :rofl:

[gurusan]

...how are you going to join copper "hairs" onto a block though. There are limitations of current machining techniques.

Also copper pins are already pretty small in the latest blocks, if they were smaller there might be a danger of them bending under higher flow rates, not to mention more viscous coolant mixtures.

[fart_plume]

There's always a problem to be resolved in any new idea..........:shrug: but is is an idea worth looking into and when I get the time I'll build a block using this idea because i want tosee how well it would perform and what the problems are.

[gurusan]

Well the biggest problem is building it. Which is a pretty large one...

[fart_plume]

well actually the best way i can think of right now would be to machine the block with a flat bottom thin as feastible without reducing strength for appying pressure to the cpu. Punch a pattern of holes in the base, insert wire bundles into the block, silver solder them in (for manufacturing a solder bath would be more efficient) lap the bottom of the block, and a tool would have to be developed to make sure the hairs were properly spread out. Never said it would be simple just and idea .................

[jas420221]

I just wanted to say how entertaining FPlumes post in this thread have been...

[GreyMouser]

Not to mention, the possible high cost of fabricating this "hairy" heatsink... Extreme performance is nice, but there are feasibility limits to everything. Evaporative cooling becomes wasteful, and high-maintenance. Exotic materials raise the cost of ownership. Of course there's a lot of buzz in the phase-change forums; refrigeration units are becoming smaller and more efficient every day. There's only so much you can do with a closed-loop cooling system. Phase-change can induce condensation. We don't, unless we're using a chilled system. People like Eddie are constantly exploring new waterblock designs; the ones that are on the market today are a far cry from the old "maze 1" blocks when things were just getting started. But there are only so many ways you can configure a closed cooling loop, and there are only so many ways that manufacturers are improving pump designs.

In terms of the whole technology curve, phase-change cooling is still in its infancy. Technology jumps are rapid and exciting. So the forums hum. Water-cooling is pretty mainstream today; changes here are going to be evolutionary, not revolutionary. We already went through the infancy stage of massive changes in technology; the move from fountain pumps to high-tech micro-pumps like the DDC has already happened in this space. So, warning someone to avoid a mixed-metals loop, and making suggestions on what particular pump or other hardware to load into their loop is where it's at these days, for better or worse. When Eddie or D-Tek or Swiftech come out with a new block or whatever, there's a buzz. It just doesn't happen as often as in the phase-change space.

-M

[fart_plume]

Here's some other things to consider in my design idea. You now have little silver plugs that are in direct contact with the ihs and in contact with the hairs, which have low mass to surface area that will heat faster because of the low mass, but due to the higher surface area ratio will remove that heat quicker. Might be negiligible might not.
So now the question of durability versus effeciency, do you use a 36ga. hair or a 22ga. hair or somewhere in the middle? larger smaller? what length of hair?

I believe with this type of block a high flow rate would in fact be benifical, perhaps much high flowrates than are used now.

[aspire.comptech]

You think the "hairs" would hold up under the pressure of the flow of water?

I have a feeling it would just wind up looking like a bad comb over.

[fart_plume]

See, this the kind of discussion needed here. that is where experimentation comes in is to find the optimal diameter of the hairs to counter the effects of flow and yet conduct heat effectively. Also what type of copper alloy? soft or hard? which would be best?

[jas420221]

I think its about time this spawned its own thread. Poor dude was asking about flow and here u are talking about cillia/flagella type cooling...!

[gurusan]

it's an interesting idea but:

a) building it will be difficult
b) if you can, it will be VERY expensive
c) it will be even more expensive to the buyer and you wouldn't sell any
d) it might get you another 5c towards ambient, max.....


The best way to improve a water setup is to get a larger radiator while attempting to maintain flow rate. Just keep the water around ambient and that's pretty darned good.

[hotdun]

I thought you were thinking outside of the box here? The new GTZ has "a pin matrix composed of 225 µm" pins. The human hair is around 180 µm so this isn't anything new.

[fart_plume]

still a pin, wire can still be drawn finer than a human hair.

[gurusan]

how is that going to withstand even a mediocre flow rate of a viscous fluid?

[eXa]

Im sorry FP...

Its really hard to take u serious when u comes with several claims wich aint even right. U comment on stuff u seeme to lack real understanding of. And when confronted and proved wrong u just push it aside with a "its a moot point"
I get this facepalm feeling.
I do understand the point u are trying to make but u are ruining the point for urself.

About the hair idea. What do u think will happend to the hairs ability to move the heat away from the base when its so thin? It looses that ability. the thinner it gets the closer to the base the hair loose its heat. So most of the hairs length is useless. The water will have taken out all the heat from the hair before its even reached halfway up its length. The rest of the hair will just be wasted material and probably just hinder flow.
Blocks like tha GTZ and the EK Supreme tries to balance that. Its no point in making the pins or fins much longer since its too thin to move the heat any further.
Swiftech and EK and others actually knows what they are doing, they have the understanding u seeme to lack, so again im sorry but its hard to take u seriously.
But as i said, i do understand the point u are trying to make and its a good one.

[fart_plume]

Who knows? something to be tested, and that's the point. Trying new things, old things, and anything to find a different solution. a smaller diameter flexible hair will withstand a vicous fliud more readily than a stiff one will, simply because it can giveandgo with the flow.........instead of trying to resist it.......

[gurusan]

a smaller diameter flexible hair will withstand a vicous fliud more readily than a stiff one will, simply because it can giveandgo with the flow.........instead of trying to resist it.......

:confused:

[eXa]

yes, Trying new stuff is good.

hair thing: U cant make a flexible har out of silver or copper, it will bend a few times then break. Try bending a wire a few times and see what happends... u get the point.
Edit: Here is a good hint for you: http://en.wikipedia.org/wiki/Fatigue_(material)

[fart_plume]

who said anything about the length of the hairs? not me. I'm just speaking of the diameter of the hairs. the length would be relative to diameter as would the depth of the block. As i said before a system to be tuned.

[eXa]

But if its just 0.1mm high it aint much of a hair...:rolleyes:
And then u only have a really fine pin grid...

[fart_plume]

yes, Trying new stuff is good.

hair thing: U cant make a flexible har out of silver or copper, it will bend a few times then break. Try bending a wire a few times and see what happends... u get the point.
Edit: Here is a good hint for you: http://en.wikipedia.org/wiki/Fatigue_(material)


Yes thats true, however if a closed system with the flow going only one direction how is the copper wire going to get bent back and forth?

[eXa]

the will be some turbulence moving the hair sligthly. Maybe it wont break imidiatly but sooner or later it will. I for one wont have copper hairs in my pump.

[fart_plume]

ok, now stop nay saying and come up with ideas to make it work........................

[eXa]

Im more of a realist and i cant say that i have the skill or knowledge to develop new waterblocks

But i tell u this, u wont see any revelutionary with normal waterblocks. there is just so much u can do with the amount of space we got. But hey, prove me wrong and i will be delighted.

[Shocker003]

See, this the kind of discussion needed here. that is where experimentation comes in is to find the optimal diameter of the hairs to counter the effects of flow and yet conduct heat effectively. Also what type of copper alloy? soft or hard? which would be best?

I like your idea, in as much as it sounds impossible to build at the moment.
I believe it will reduce a chip´s temp by some good °c:up:


Who knows? something to be tested, and that's the point. Trying new things, old things, and anything to find a different solution. a smaller diameter flexible hair will withstand a vicous fliud more readily than a stiff one will, simply because it can giveandgo with the flow.........instead of trying to resist it.......

With this theory of it been flexible it woun´t work. When your copper hairs are been pushed to one direction from the flow, they will loss their form(get compacted and will act like a plain surface with few contures). Your aim will be lost and i presume you need a metal that woun´t be totally flexible and can regain their original form when the pump is turned off.



Yes thats true, however if a closed system with the flow going only one direction how is the copper wire going to get bent back and forth?

Think towards this direction, the block will be very restrictive and if the block will have more pins or hair like pins it will have more surface area that come in contact with the coolant. I strongly believe it´s possible, but not with flexible hair. Think of a block in the form of a cave with stalagmites and stalagtites(pins) all over the inner chamber of this block and the water has to be channeled towards a direction so that it will come in contact with all the pins( it will be very restrictive and effective in removing heat). The whole block will be removing the heat and not just the copper base. Heat moves towards colder region, it will be spread all over this block and not just concentrating at the base of the block like most traditional water blocks

[fart_plume]

Now this is what I've been trying to do is to get people thinking and not just reciting the same stuff over and over. To try different things, help out by coming up with a solution not by telling someone to spend more money on bling. We have one thinker how many more of you are there?

[hotdun]

Now this is what I've been trying to do is to get people thinking and not just reciting the same stuff over and over. To try different things, help out by coming up with a solution not by telling someone to spend more money on bling. We have one thinker how many more of you are there?

Start a new thread hijacker.....

[fart_plume]

Start a new thread hijacker.....

True I did hijack this thread and to the op I apologize, @ hotdun Mind your manners...........

[NaeKuh]

I like your idea, in as much as it sounds impossible to build at the moment.
I believe it will reduce a chip´s temp by some good °c:up:


how would having pins on the top help?

i think you guys missed my point. Surface coverage is not very dependant inside a block. Otherwise scott's steping would of been worse then a bow.

[fart_plume]

I got that , I just didn't want to shoot down someone who's trying to come up with an idea, not yet anyway, too early for that just want to get the thinking to start.

Again sorry to the op for the thread jack!

[Shocker003]

how would having pins on the top help?

i think you guys missed my point. Surface coverage is not very dependant inside a block. Otherwise scott's steping would of been worse then a bow.

When the whole block is made of copper and the inner chamber is filled up with these pins. That will mean that the heat conducted from the cpu his will spread to all parts of this block. It´s not just a normal block with copper base and acyl or Del top. Meanwhile reread my post, No flexible pins as proposed by Fart and it´s just like your tradition block but using the same base as it´s side walls and top. I believe Fart has a point when he said we should think biologically. Just think of a stomach with all these filaments lining it´s wall that help to absorb the digested food into our blood streams. But in this block reverse is the case. In heatsinks with heatpipes, the pipes help to move the heat from the base to the fins attached on them(you know the rest of the process). The heat will equally be conducted all over this block and the coolant will come in contact with more heat dispating surfaces(pin lined walls).

NB: i know my english sucks:p: Please try to over look my mistakes or correct me;) Meanwhile explain scott's stepping to noob please:shrug:

[hotdun]

True I did hijack this thread and to the op I apologize, @ hotdun Mind your manners...........

Sorry, I'm just not a fan of the whole "holier than thou" attitude. I personally like the way this forum's going and get alot out of it. If nothing ever changed, I'd still be happy....

Coming in outta the blue and saying that everyone is being complacent and then saying you have new ideas and that you're from a race of super engineers and animal cooling and then never really telling people what your ideas are.....all in someone elses thread..... I don't like that approach...

[aspire.comptech]

When the whole block is made of copper and the inner chamber is filled up with these pins. That will mean that the heat conducted from the cpu his will spread to all parts of this block. It´s not just a normal block with copper base and acyl or Del top. Meanwhile reread my post, No flexible pins as proposed by Fart and it´s just like your tradition block but using the same base as it´s side walls and top. I believe Fart has a point when he said we should think biologically. Just think of a stomach with all these filaments lining it´s wall that help to absorb the digested food into our blood streams. But in this block reverse is the case. In heatsinks with heatpipes, the pipes help to move the heat from the base to the fins attached on them(you know the rest of the process). The heat will equally be conducted all over this block and the coolant will come in contact with more heat dispating surfaces(pin lined walls).

NB: i know my english sucks:p: Please try to over look my mistakes or correct me;) Meanwhile explain scott's stepping to noob please:shrug:

The problem is, the heat will never get a chnace to spread to the top half of the block.

A. The block cannot be one solid piece of metal. It would likely need to be cut in half. Heat would not readily move from one half to the other.

B. A block never gets hot to the touch as it is. Feel the top of an Apogee GTX. Unless your really saturating your radiator it will always be cool to the touch.

C. Imagine how much more it would cost. Instead of buying one water block, you'd essentially be buying two.

[gurusan]

Also we are pretty close to the maximum efficiency you can get out of these setups anyway...the latest blocks are pretty damned good. Even if you made a Dtek Fuzion out of diamond I don't think it would do much more than 5-8c better. Actually you can figure it out but my brain hurts too much at the moment and I don't know where my material properties handbook is....not to mention I forgot how. damned summer vacation is too long

[Levish]

Been a while since we had this much drama :D

For a answer to the OP, I haven't seen such a point reached where more flow resulted in worse performance but everything with current waterblocks that had charts posted had strong diminishing returns over 1.5gpm and nearly no worthwhile gains at all over 2gpm.

[Shocker003]

The problem is, the heat will never get a chnace to spread to the top half of the block.

A. The block cannot be one solid piece of metal. It would likely need to be cut in half. Heat would not readily move from one half to the other.

B. A block never gets hot to the touch as it is. Feel the top of an Apogee GTX. Unless your really saturating your radiator it will always be cool to the touch.

C. Imagine how much more it would cost. Instead of buying one water block, you'd essentially be buying two.

I never intended that it should be one solid piece of metal. It will be in two or three parts that can be screwed together.
What makes you feel the heat woun´t be conducted to other parts of this block. As far as the parts that make up the said block are in contact, the heat will spread. Take note that heat always moves towards colder region.
Actually i don´t own an Apogee GTX and i strongly believe if a company like D-Tek launches such a block, most people will pick it up, as far as it blings and could drop some °c. I don´t believe it will be more expensive than these
http://www.aquatuning.de/shopping_cart.php/products_id/4076

[Shocker003]

Also we are pretty close to the maximum efficiency you can get out of these setups anyway...the latest blocks are pretty damned good. Even if you made a Dtek Fuzion out of diamond I don't think it would do much more than 5-8c better. Actually you can figure it out but my brain hurts too much at the moment and I don't know where my material properties handbook is....not to mention I forgot how. damned summer vacation is too long

Well most people wcing and ocing will be happy if they could lose 5-8°c from their cpu temp. When i have 24°c in my computer room, my core0 always idle @30-32°c minus 5°c like you wrote=25-27°c. Which most people will brag about and take note we are not anywhere close to maximum efficiency as far as watercooling is question.

[gurusan]

erm yeah 5c is nice, but I said if it were made of diamond. Do you have any idea how much that would cost?

[Shocker003]

erm yeah 5c is nice, but I said if it were made of diamond. Do you have any idea how much that would cost?

Lets be realistic man. Who will even waste some diamonds on water block except if Bill Gates or the Queen wants a pimped PC;)

[eXa]

Well most people wcing and ocing will be happy if they could lose 5-8°c from their cpu temp. When i have 24°c in my computer room, my core0 always idle @30-32°c minus 5°c like you wrote=25-27°c. Which most people will brag about and take note we are not anywhere close to maximum efficiency as far as watercooling is question.

Idle? Who gives a s*** about idle? Its load that counts.

[Shocker003]

Idle? Who gives a s*** about idle? Its load that counts.

Well with four threads of prime95 running i can only hit 46°c-48°c with room temp @ 24°c(I hope you are satisfied)

[eXa]

Thats low... more volts and more speed then :p:

[gillbot]

hey FP, i hate to throw a wrench at ya.

But you cant get evap occuring inside a closed loop, unless its due to cavitation. Which then its actually vapor stuck in your inlet.
:T

Your idea is good, if the waterblock was open and you could evap as it traveled, but in a closed loop, there is no such thing as evap, unless your talking about a bong cooler which requires an exhaust.

(sub note: the tubes are actually pourous, so you do lose coolant over time due to evap, and osmosis but not on a degree where it would cool the liquid)

Your idea of increasing surface area is good, however thats easily accomplished via other methods like glass beading, and cross slits or drilling.

You cant get evap occuring unless your on HWLabs RxN which never got launched.
Im guessing due to coolant loss and humidity.

You missed the mark naekuh....

Cavitation is caused by ENTRAINED air suspended in the liquid. Cavitation is the expansion and subsequent collapse of this entrained air and it usually occurs on the trailing edges of the impeller vanes, not at the inlet.

Remember the soda bottle metaphor? Cavitation is like shaking a soda bottle, immediately releasing all that pressure then forcing it completely back into the bottle without any spillage, then closing the lid.

[gurusan]

Lets be realistic man. Who will even waste some diamonds on water block except if Bill Gates or the Queen wants a pimped PC;)

that's my entire point. Realistically there are not many more gains to be had....

[DoGMaN]

What happened to my poor thread :shakes:

[NaeKuh]

You missed the mark naekuh....

Cavitation is caused by ENTRAINED air suspended in the liquid. Cavitation is the expansion and subsequent collapse of this entrained air and it usually occurs on the trailing edges of the impeller vanes, not at the inlet.

Remember the soda bottle metaphor? Cavitation is like shaking a soda bottle, immediately releasing all that pressure then forcing it completely back into the bottle without any spillage, then closing the lid.

YES SIR! :rofl:


yeah what my friend gill said!


Too much flow:
http://www.tkg.tv/images/Another%20Blue%20Sky%20in%20Niagara%20Falls.jpg


What happened to my poor thread :shakes:

Cliff notes:
- Your pump wont give you super head pressure or flow, so you need not worry.

For reference see scott's post. Trust him.

I hate to be the one to tell you this but the 1260 is an awful pump for water cooling...

[gillbot]

Sadly, Niagara has not been at full flow for many years....


The Falls flow is further halved at night, and during the low tourist season in the winter, remains a flat 50,000 cubic feet per second (1,416 m³/s). Water diversion is regulated by the 1950 Niagara Treaty and is administered by the International Niagara Board of Control (IJC).

They have been limiting the flow at Niagara Falls for a long time, diverting it through the power station. There was a petition online for a while asking for a national Niagara Day where they let them go at full flow at least once for all of this generation to see.

[nikhsub1]

Sorry for OT, but gillbot your sig is pretty rude dude. You may want to remove the 'STFU' bit.

[NaeKuh]

Sorry for OT, but gillbot your sig is pretty rude dude. You may want to remove the 'STFU' bit.

nah scott its alright, thanks for the thought.

Gillbot is a buddy of mine. Its a joke cuz i keep making fun of his noobish posts sometimes. :rofl:

[nikhsub1]

nah scott its alright, thanks for the thought.

Gillbot is a buddy of mine. Its a joke cuz i keep making fun of his noobish posts sometimes. :rofl:
Alright carry on then :welcome: