in my fight against stupidity, I need some help here....

heres the quote of the day....

"According to physics, air is better for disippating heat. Water has a very low transfer rate compared to most materials. The big difference is that when you have cooling with water you have a liquid that has lower heat transfer than air but the bulk of it and the circulation changes all that. What makes water cooling better is that it flows over, and the water is then cooled. Your cpu technically is still air cooled, you're just creating a much much larger surface area for the air to cool it by using a conduit for heat transfer that is water cooling. Instead of cooling just where that little chip is you're cooling all around an entire radiator or tubing complex(whatever word you wanna call it). There are 10 times better conductors of heat one could use than water, air just being one of them but you can't really circulate and cool the air with readily available materials. Water is just available in large quantities so that's used as the conductor. Most of the better conductors also evaporate as well. The really serious people start getting into non-electrolytic fluids and dry ice, because honestly, water sucks. It's better than heat-sink/fan combo but in reality it's basically just giving the air more area to cool, the surface area works very well in cooling it then.

Air is more efficient than water when given the proper surface area. liquid cooling is required at the moment to give that bigger surface area so they are used together. If you could isolate the water in something like a thermos so the air can't really cool it you'd quickly see many problems after the water has passed through a couple times.

Another good point about water, it won't get too hot! it takes a lot of energy to heat up water since it's a fairly bad one! That's a big plus actually! just circulate it fast enough to keep it able to absorb more and more and you're fine"


now I know I have seen graph and statement after statement about water is 4 times better then water, and direct die is ten times better then water...

just I cannot remember where I heard this.
help me find the right proof here to get this right, without it just becoming a flame war arguement.

ROFL That's a good one

Let's start with the quote

According to physics, air is better for disippating heat. Water has a very low transfer rate compared to most materials. The big difference is that when you have cooling with water you have a liquid that has lower heat transfer than air

Click to expand...

I'd like to see those physics. Water does indeed have a low "thermal transfer rate" but this is for CONDUCTION in a non moving sample only. Comparing air and water under the same conditions i.e. not moving means that water will come out on top, but both are still crummy compared to a metal and even worse when compared to copper. So, in effect that statement is wrong.

Some numbers for ya:
Conductivity measured in W m-1 k-1
Air: 0.026
Water: 0.58
Aluminium: 210

There are 10 times better conductors of heat one could use than water, air just being one of them

Click to expand...

Oops, wrong again. this implies that air conducts heat ten times better, where in fact water conducts close to 20 times better.

Water cooling is indeed a more elaborate method of spreading heat, as heat radiated and convects away from all the hot bits, but the key is the radiator. This has a massive surface area so the crummy conductivity of air is compensated for. Plus the straight line conductivity is more or less negated due to the turbulence and mixing.

Air is more efficient than water when given the proper surface area. liquid cooling is required at the moment to give that bigger surface area so they are used together. If you could isolate the water in something like a thermos so the air can't really cool it you'd quickly see many problems after the water has passed through a couple times.

Click to expand...

The only reason heatsinks using air work is because of the large fan sitting on top. Ever tried to run a PC without a fan? If you're lucky you have overheating protection. This statement implies that having a large surface area means that air will beat water. But by having no airflow. I wouldn't like to see that being tried (Yes I know there are passive HS for Opterons, but airflow in the server case counts as a fan). I expect if you isolated an air cooled CPU in a thermos it would encounter problems too.

Another good point about water, it won't get too hot! it takes a lot of energy to heat up water since it's a fairly bad one! That's a big plus actually! just circulate it fast enough to keep it able to absorb more and more and you're fine

Click to expand...

Ok this is a bit confusing. This talks about specific heat capacity, and in this department WC pwns Air. The specific heat capacity of the substance is defined as the amount of energy required to raise 1 gramme by 1 degree C.

Numbers again: (J kg-1 K-1)
Water: 4186
Air: 1000 (100C)
Aluminium: 900
Copper: 390

These numbers are a bit odd because they are in fact for mass. 1 gramme of water will be a lot smaller than 1 gramme of air, and aluminium will be even smaller than that. Even so, water's very high specific heat capacity makes it very suitable for coolant. It takes a lot of energy to heat it up (I promise you I understand the physics behind this) and so it can remove a lot of energy per unit mass. It is true that more flow rate is better, because of thermodynamics and all that baloney. Read the sticky on it for really nitty gritty detail.

Sorry for bashing your quote but it was a bit incorrect.

Credit to sources: http://www.monachos.gr/en/resources/Thermo/specific_heat.asp

Part of that quote is true. -I feel stupid for even saying that, so much of it makes no sense-

The water in a water cooling setup is just a conduit for heat. You are moving the heat from the CPU to the radiator where it is dissipated into the air.

It seems to be saying that if you had a heatsink that had the same amount of surface area as the radiator in a watercooling system you would achieve better temps. This sounds like it could be right. You may need more airflow with the giant heatsink, but it should achieve better temps. This is because in watercooling you have 2 additional convective heat transfers, one from the waterblock material to the water, and another from the water to the radiator and then from the radiator to the air. Where air cooling uses only one such transfer, from the heatsink to the air. This is assuming you have great case ventilation, or the air going through the heatsink will be too hot to effectively cool. It is quite easy to get outside (cool) air into a radiator, much easier than a heatsink which is stuck in the middle of the case.

We use watercooling for a number of reasons.

1) A copper heatsink with as much surface area as a radiator, even a small 120mm radiator would weigh much too much to be safely attached to the socket or even the motherboard.

2) To achieve good temps with realistically sized copper heatsink you need so much airflow that the noise is deafening. Watercooling is much quieter.

-- I'm sure there are many more, which I'm forgetting.


Now - the part I have trouble with is saying water has a low transfer rate.

Water is much better at convective heat transfer than air.

Fluid, conditions: Heat transfer coefficient, W/m2•K

Air, natural convection: 5
Air, forced convection :10 - 100

Water, natural convection : 200
Water, forced convection : 500 - 10,000

from: http://chemeng1.kat.lth.se/staff/ulf_b/hc_heat.htm#top

We are talking about forced convection for both water and air.

The quoted seems to be assuming that because the temperature of air will increase more than water that the air is in fact absorbing more heat. This is not true - and gets to the concept of specific heat. I hate re-typing things, so go here: http://ch185.semo.edu/specheat/enter.html it is a very good explaination.

Think about what happens when a hot metal is held in the air – it takes quite a while to cool. Now dip that same material into a pool of water and it cools at a very fast rate even if the water is the same temperature as the air. The water is able to absorb heat much faster – and its temperature change is very low.

Also - when considering a RATE in heat transfer you are not only referring to material properties, but also the geometry of the apparatus and other system parameters such as flow rate. So to make a blanket statement that water has a low rate of heat transfer is completely invalid. I could easily design a specific situation where the heat transfer rate for a water system would be far more than air in a dissimilar system, and vise versa.

The example of the water in a thermos is absurd because air in the same situation would do the exact same thing – in fact ANYTHING would heat up and lose it’s ability to cool in that situation…pesky laws of thermodynamics.

It also comments on dry-ice, which is great, but only for a short period of time. Apply a heat source to dry ice and it will quickly sublime. Dry-ice and liquid nitrogen are good for short term super overclocks because they exist at such a low temperature (high delta T equals a high rate of heat transfer). But never long term because they quickly return to their gaseous form when heated.

This thread could probably be in the technical forum.


Edit: Where did that quote come from?

The whole thing is ridiculous as noted, but I still like the summary where he writes, "water is 4 times better then water" which actually makes about as much sense as the rest. True it's pointless to attack that point but to me this is the point where my arguement becomes. "You don't seem to be right on any of your facts. Most are the opposite of your claims. I'd be interested to see support for any of it."

Edit: Where did that quote come from?

it came from another forum where I was talking about overclocking, and someone started trying to tell everyone that air cooling was better then water...

haha

you have to remember that a radiator is going to do a much better job with heat then a hs, according to the small engine book i used in high school water disipates heat 2.5 times better then air, what is why most car engines are water cooled and not air cooled

This guide will explain some tidbits and some terminology that is often confused in these discussions. I pulled out some of the most relevant-to-the-thread quotes, but the entire guide is really worth reading and there are some other very useful equations in the article also:

General Heat Transfer Guide:
http://www.amdmb.com/article-display.php?ArticleID=105&PageID=1

Conduction: Heat transfer through a solid material. A heat sink conducts heat from the die to its fins/pins. Conduction increases with increasing temperature differential, increasing conduction coefficient, increasing cross-sectional area, and decreasing material thickness.

Conduction Coefficient: A measure of how efficiently a solid conducts heat. Among all materials, diamonds have the best conduction coefficient. Among materials commonly used in electronics, the top conductors are silver, copper, and gold. Aluminum is also a respectable conductor.

Convection: Heat transfer from a solid into a liquid or gas. The energy transferred through the heat sink leaves via convection to air or water. Convection increases with increasing temperature differential, increasing surface area, and increasing convection coefficient.

Water and air both cool solids via convection. The primary difference is the convection coefficient of water compared to air. At the same conditions (temperature, flow velocity, etc.), water is on the order of twenty times more efficient at convection than air is. Air holds about 1000 joules of energy for every degree Celsius that a kilogram of air changes temperature. Water holds about 4000 joules for every degree in a kilogram. Liquid water is about 1000 times as dense as air. This means a smaller mass and much, much smaller volume of water is required to contain the same energy as air.

Click to expand...