I don't get all this delta vrs flow vrs moon phases stuff, but can you explain to me what you told GBN about deminished returns with his water cooling setup? Not pissed just want to understand it. On my system as i have tinkerd with it..the bigger and faster the fan the better it cooled. I want to understand i really do. Thanks bud. and maybe after the holidays have passed i can start milling that chunk of Silver I got.

no answer m8? or just didn't see it? BTTT

Quote: Originally Posted by FizzledFiend no answer m8? or just didn't see it? BTTT

I just saw it...what I told GPN and several times in the forums is there is a point where more airflow doesn't cool any better. as a matter of fact if you try to force air through the radiator, there is a point where the turblence actually causes less air to make it's way through the fins.
The smoother air flow over the fins is the most efficient, over blowing them does nothing for temps.

tDelta is simply the difference between ambient and the temp of whatever you are measuring.
I have recorded a tDelta of 2-3C in the coolant using a standard fitting 120x38mm delta 130cfm fan on 7volts at aprox 60-70 cfm. Blowing 200+CFM through that same radiator will yeild no better coolant temps but will make you deaf.

contrary to incorrect theorys there isnt a wind chill factor that occurs in PC H20 cooling and you will never reach at or below ambient temps, period.

Quote: Originally Posted by ZENNZZO I just saw it...what I told GPN and several times in the forums is there is a point where more airflow doesn't cool any better. as a matter of fact if you try to force air through the radiator, there is a point where the turblence actually causes less air to make it's way through the fins. The smoother air flow over the fins is the most efficient, over blowing them does nothing for temps.

yep yep... go read up on some fluid dynamics text books... specifically flow between flat plates(aka fins) the optimal air speed can be found at the speed where the air between each adjacent fin just becomes turbulent from previous laminar flow.... college level stuff have fun reading

Quote: tDelta is simply the difference between ambient and the temp of whatever you are measuring. I have recorded a tDelta of 2-3C in the coolant using a standard fitting 120x38mm delta 130cfm fan on 7volts at aprox 60-70 cfm. Blowing 200+CFM through that same radiator will yeild no better coolant temps but will make you deaf.

dun fully agree with this... it depends on rad and and fin spacing really.... agree with the deaf path tho

Quote: contrary to incorrect theorys there isnt a wind chill factor that occurs in PC H20 cooling and you will never reach at or below ambient temps, period.

try evap with a bong or volenti... you CAN reach ambient or below ambient if you do it properly and the environment conditions are good... just don't try it in the mid of winter without a heater nearby then you might have ice-cooling

Quote: Originally Posted by ZENNZZO I just saw it...what I told GPN and several times in the forums is there is a point where more airflow doesn't cool any better. as a matter of fact if you try to force air through the radiator, there is a point where the turblence actually causes less air to make it's way through the fins. The smoother air flow over the fins is the most efficient, over blowing them does nothing for temps. tDelta is simply the difference between ambient and the temp of whatever you are measuring. I have recorded a tDelta of 2-3C in the coolant using a standard fitting 120x38mm delta 130cfm fan on 7volts at aprox 60-70 cfm. Blowing 200+CFM through that same radiator will yeild no better coolant temps but will make you deaf. contrary to incorrect theorys there isnt a wind chill factor that occurs in PC H20 cooling and you will never reach at or below ambient temps, period.

Also, the faster a fan moves the more turbulence the blades make for each adjacent blade making them less efficient , meaning at some point a fan will actually move less air the faster it goes.


EDIT: OOPS, just read duhs post, he said the same thing.

ANOTHER EDIT: BTW duh, before you start disagreeing with Zen on cooling issues, you may want to prove your worth in the field, he certainly has, and he knows his shizzle.

Quote: Originally Posted by duh_bing yep yep... go read up on some fluid dynamics text books... specifically flow between flat plates(aka fins) the optimal air speed can be found at the speed where the air between each adjacent fin just becomes turbulent from previous laminar flow.... college level stuff have fun reading dun fully agree with this... it depends on rad and and fin spacing really.... agree with the deaf path tho try evap with a bong or volenti... you CAN reach ambient or below ambient if you do it properly and the environment conditions are good... just don't try it in the mid of winter without a heater nearby then you might have ice-cooling

If a dude is asking questions about how to understand basic watercooling, I really doubt that advanced stuff like bong cooling is relevant. You want to talk phase change? Ask Zenn.

I appreciate your interest in this thread, but am still trying to decide where you helped.

Quote: Originally Posted by duh_bing yep yep... go read up on some fluid dynamics text books... specifically flow between flat plates(aka fins) the optimal air speed can be found at the speed where the air between each adjacent fin just becomes turbulent from previous laminar flow.... college level stuff have fun reading dun fully agree with this... it depends on rad and and fin spacing really.... agree with the deaf path tho

Is it the definition of tDelta you don't agree with, or the fact that My BIXll runs cooler at the approx CFM listed at 7vts. rather than the max of 12vts and 520CFM powered by the fans?

Quote: Originally Posted by duh_bing try evap with a bong or volenti... you CAN reach ambient or below ambient if you do it properly and the environment conditions are good... just don't try it in the mid of winter without a heater nearby then you might have ice-cooling

Well now we arn't talking PC H20 cooling now are we, sounds like we are talking PC evaporative cooling HUH? There are a few examples out there but not really enough to even call it mainstream huh?
There are also folks who use the earth as their radiator. 18-24" below the frost line...the ground stays within 10c year round and a burried grid of copper pipes serves to cool the coolant...
however...
we are talking about a large big cfm fan that was applied to a heater core to be mounted inside the front of a midtower...
I have personally tried the 110vt version of that fan and like I said in the post above I got better coolant temps when I used a smaller cfm fan that fit the core better and made smoother air over the fins ....

Who are you anyway? and what are you using to cool with?

As far as the bong/evap coolers go. to reach ambient with anything more than an old Duron in the loop you're going to need a lot of water and a whole helluvalot of surface area to dissipate that heat. Reaching ambient tempaturers is far more easily done with a radiator (IMO).

Also are we talking CPU @ load = ambient or coolant temp = ambient? Because unless I'm missing something once the coolant temp is at ambient its all up to the waterblock what temp the CPU will be @ load and it seems that here we're talking about radiators and thus coolant temp.

Quote: Originally Posted by Arkangyl As far as the bong/evap coolers go. to reach ambient with anything more than an old Duron in the loop you're going to need a lot of water and a whole helluvalot of surface area to dissipate that heat. Reaching ambient tempaturers is far more easily done with a radiator (IMO). Also are we talking CPU @ load = ambient or coolant temp = ambient? Because unless I'm missing something once the coolant temp is at ambient its all up to the waterblock what temp the CPU will be @ load and it seems that here we're talking about radiators and thus coolant temp.

Ark,
ambient temp is the temp of the surounding atmosphere...like room temp?
It is the temp of the air that is used to cool with...unless you do any type of evaporative cooling you will not get below ambient with straight air...
Coolant temp without load at best will be ambient, as soon as you apply the load of a CPU or GPU you can add at least 10-12C to the temp of the coolant, most systems will go like 15-25C above ambient.

seems to me that duh is agreeing with zennzo. he pretty much said to get below ambient temps then you need to go to phase change.

all other points duh is just pointing fizzle where to do some further reading on the subject

Apologies if i sounded like some bigass-know-it-all started visiting pr only recently

Luna mod: not disagreeing with zen completely... just pointing out that theres alt. water(as coolant)cooling methods where below abient can be achieved

zen: erm... din see a bix mentioned so i assumed you are talking abt a hypothetical rad... how well a rad cools w a particular airflow speed depends on the spacing of the fins..

i regard bong/volenti cooling as water cooling also, phase change i look more at refrigeration type using all the funny materials (R134 etc etc) sorry if there is a misunderstanding...

below ambient i was refering the the coolant(water), wrote in the first time i posted which got swollowed up by the net and i had to retype the entire thingy...

Enko: thanx for helping me re-interpret stuff

ok did a few late night sesions reading about this stuff..to put in terms I can understand, and so maybe some others. It appears that @ a certain point there is MORE air than there is space to move..so it starts backing up and the effect gets worse and worse toill it caps off. Thismake total sense now. I haven't tried too many different rads and always went with huge mothers of radiators. so there must have been the reason for the higher I went the better it got..I never had more air than space to move it. Thanks Zen. now to the delta part if I understand it your saying that delta is the # difference between surrounding air and the heat source? thats simple enough if i have it right..

on a side note, Zen you got the addy to where you got that chunk of acrylic? I think to save time and money I am going to make the body out of acry and the base, thus cpu contactpoint, out of that silver.
thanks Zen, and to the rest of you guys for chiming in on this.

It was mentioned about buring passive radiators below the frost line..thats what I have hear and have reduced my noise considerably. I do use about a 2 gal res and have noticed some fluctuations. ic an only assume that here is where thats happening. I am going to insulate it and see if it stabilizes.

I did a little research and discovered that the higher the pressure, the lower the freezing point of water. Which is opposite almost every other element or compound.

1 atmosphere or 14.7 psi, water freezes at 0 degrees celcius

1000 atmospheres or 14,696 psi, water freezes at approximately -5

There are fittings and materials that can handle 16,000+ psi, but i wouldn't want to be anywhere near my pc when it fails.

hmm thats interesting.....is there a difference between air pressure and watter pressure? lets say i pressurize my cooling loop with air instead of water, could i then excert the same force of pressure with water and get the same results? This is how we test vehicle radiators, pressurize with air and if it holds air pressure it will hold the same water pressure..I have to assume it would be the same for our cooling loops as well...seems the be the same principle aye?

Ok now about the reverse? what if i created a vaccume in my cooling loop...does the freezing point increase? or would it boild sooner? What effects would pressure and vaccume have on the cooling properties of water?

don't really think pressurized water and pressurized air can compare. water can't compress while air can so once you fill up a container with water there's no way to increase the pressure on the water. if you forced more air into the system, you'll just be stressing a fitting that will threaten to blow.

Quote: Originally Posted by FizzledFiend hmm thats interesting.....is there a difference between air pressure and watter pressure? lets say i pressurize my cooling loop with air instead of water, could i then excert the same force of pressure with water and get the same results? This is how we test vehicle radiators, pressurize with air and if it holds air pressure it will hold the same water pressure..I have to assume it would be the same for our cooling loops as well...seems the be the same principle aye? Ok now about the reverse? what if i created a vaccume in my cooling loop...does the freezing point increase? or would it boild sooner? What effects would pressure and vaccume have on the cooling properties of water?

no...
First rule of water cooling ...WATER in itself does not cool....it is only the vechicle than can transfer energy on it's surface...
unlike automotive cooling where our main objective is to keep from boiling over, and a very high operating temp, with high pressure and mass flow. As well as a large area that the energy is being crated from.

we cool a cpu off, that has at the most 165 watts, that radiates that energy from an area the size of a postage stamp, and smaller.
We run circulation pumps that create adequate coolant flow at a low pressure, under 10PSI.
we strive for the smoothest flow, with the least restrictions.
At the energy source, the CPU, we have found that "IMPINGMENT" style blocks, provide the much needed tubulence at a point, where the most surface area of the block is exposed....the channels.

Now we rely on a smooth flow to the next stage of the cooling process where we have coolant that is energy filled, dump that energy into the surounding ambient by way of a radiator and fans.

we have found that the best balance of coolant flow and Air flow over the fins of the radiator, is what will rid the coolant of the CPU energy, so it can circulate back to do more of the same....

Adding pressure, vaccuum etc. for that to work it has to be done with a different coolant....for it to be efficient it has to have refrigeration properties. When dealing with those extrem pressures and vaccuum, the "Phase" of the coolant is going to shift, and now you have a different style of cooling system all together. I'm sure you recognize the term...


take everything you might understand about automotive water cooling, and leave it in the glove box.. the basic lineup of cooling components is right about where any simularities end...

Hope that clears up any theory on the pricipals of PC liquid cooling...
At this point in the evolution of CPU cooling, Air coolers are able to get close, adequate even, the numbers can even be reported as close...
where they lack is in the ability to keep CPU temps steady without alot of spikey results...
I have found there is more stability in my O/C's when the temp line remains as flat as possible.
Stability means higher clocks.

The reason for cryo systems is that they can be tunned for a specific load,
They can manage the extra heat that comes from xtremely high voltages, needed
to run out of spec at insane speeds. Doing so at a very flat temp line....

For the most part, zen is right, with a few minor corrections/adjustments.

First off: you CAN get below ambient with air/water cooling. It depends on where the heat exchange occurs. As an example, the water can be cooling (either with air, or with a refrigerant such as R-134A) and then in turn, that water can be used to lower the air temperature. Think of how air conditioners work. Only exception is that they can't use water is the cooling fluid because typically, in the compressor phase, the working fluid would be well below the freezing point of water (it's not uncommon to find frozen condensate on the pipes near the compressor) and that is an example of how on a hot summer's day, the outside air is much higher than the inside ambient.

I was originally thinking that it wouldn't have been able to cool the inside to below that of where the heat exchange occurs (heat rejection by the gas-vapour refrigeration cycle), but that isn't the case. In fact, all refrigeration cycles pretty much function on that, and the difference is the amount of energy (work) required to cool it back down (a.k.a the large fan that you see outside of homes with centralized air conditioning). Possible? Yes. Efficient? not really. Expensive? Depends on where you live.

In the case of air/water cooling, you can drop below ambient, but it gets really hard. Water isn't entirely incompressible, but most of the calculations are done on the assumption that since water is only something like 99.999973% compressible, that for all practical intents and purposes, it an incompressible fluid. That's also done to make the math easier, though there are programs now that can handle compressible Newtonian fluids without any difficulty. It may take a while, but it'll do it.

With regards to the whole issues of fins, fin spacing, turbulent vs. laminar flow; laminar ISN'T always better. You do want SOME degree of turbulent flow because that is what would help in the heat exchange process. The only caveat to that is the degree of turbulence. Like anything, too much of it isn't good. For me, my typical rule of thumb for heat exchange with air through a heatsink/heatsink-fan assembly is to have a Reynolds number (Re) between 5000-10000. If the fins are very thin (which is the good thing, but in it comes inherent cost and manufacturibility issues), it isn't likely to resist flow, and that it would allow for a higher density of fins within a given space - to achieve 5000<= Re <= 10000, it isn't hard.

For water, through a heat block, (either dissipating heat, Qout, or removing heat, Qin), my usual rule-of-thumb for that is 2000<=Re<=4000. The conduits are usually larger in area therefore; there's some degree of turbulence, but not as dramatic as it would on the HS/HSF.

I strongly urge those that have the opportunity to take a fluids/thermo class if they can, when they can, because it can dramatically improve your ability to overclock your computer/computer components. You CAN pick up a text book, but what I find most of the time when people do that is they get lost in it because of all the variables and the concepts and equations. To have someone teach it to you, is far better than reading about it, unless you have a knack for it. In addition to that, having someone teach it to you also means that if you have questions, HOPEFULLY your prof would be competent enough in the subject to be able to answer the questions that you may have.

A small, well designed heatsink/fan is better than a large, grossly inefficient one. Likewise for any fluid cooling (gaseous or not).

Personally, I'm not a big fan of water cooling because if any of the connections/seals break at or near the comp, you're pretty screwed. Normally, I prefer a non-conductive oil (such as mineral oil) that way, while it may "feel" messier, I know that I would stand a much better chance should things fail. To date though, all of my cooling is done via stock HSFs air/air exchange, though in the winter, I typically fair better because I can drop my room temperature to ~15 C or less.

On dual MP1800, the processors usually sit around 46 C respectively, though I have gotten it as far as 41 C, but then my room is pretty much freezing (somewhere between 5-10 C ambient.)

EDITED.

Sorry, I was going to chastise a couple of you for getting into technical debates that were getting off-topic.

Then I re-read the biginning and found that the whole thing is a technical question and not a request for help on a specific problem

As I get older, I get confused easier

snip

Quote: Originally Posted by alpha754293 With regards to the whole issues of fins, fin spacing, turbulent vs. laminar flow; laminar ISN'T always better. You do want SOME degree of turbulent flow because that is what would help in the heat exchange process. The only caveat to that is the degree of turbulence. Like anything, too much of it isn't good. For me, my typical rule of thumb for heat exchange with air through a heatsink/heatsink-fan assembly is to have a Reynolds number (Re) between 5000-10000. If the fins are very thin (which is the good thing, but in it comes inherent cost and manufacturibility issues), it isn't likely to resist flow, and that it would allow for a higher density of fins within a given space - to achieve 5000<= Re <= 10000, it isn't hard.

If you examine the majority of industrial computer cooling solutions, the Reynolds numbers typically range between 7800 and 8900. Guidelines for "roughness" are that texturing between 2.5x10^-8 mm and 1.0x10^-6 mm provide optimal turbulence for heatsink fin spacings of between 3mm and 8mm. A key characteristic in designing heatsinks is (bordering upon restating what has been said above) disruption of laminar flow in a controlled manner. The degree of disruption varies, however it must happen for the cooling solution to be effective. I could go on listing other guideline specifications that I have in the numerous books on electronic cooling solutions that I have on my shelves, however I'm certain I'd bore most anybody reading it.

Lots of good points made in this thread.

One thing I didn't see, though perhaps I skimmed past it.

The fan's pressure rate. The higher it is the better. And the higher it is the higher the volume of air flow through the rad you get with it.

The new Delta Tri-Blade desing is a very good example of this. The one fan mentioned above is the Delat 130cfm @ .42" of pressure. But the Tri-Blade Delta produces .51" of pressure. So the Tri-blade Delta fan is one of the best out now for a extremly low water temps.

Anyone ever used a 2-192 heater core? I'll gain you much lower temps when at load than almost any other core. Combine this core with the high pressure Tri-blades

Fans are often the most overlooked component of a water rig.