[Alchemy]

Originally posted by Graystar
Please correct my impression if it is wrong, but I get the impression that you think the radiator doesn't need to cool as much because the water temperature is lower. This, I'm afraid, is completely wrong. Yes, the water is cooler, but the rad has to cool more water for a given time period than with a slower flow rate. So the radiator really does have more energy to transfer. It must, as that is the only way to reduce the temperature of the CPU.

I'll be sure to call up Newton, Fourier, Bouissinesq, Grashof, Brinkman, Nusselt, Prandtl, Reynolds, Biot, Sieder, and Tate and let them know their concepts of heat transfer are flawed. They will be most displeased.

When we increase flow, the starting water temperature is assumed to stay the same.

All temperatures depend on the heat transfer coefficients. Increasing one of them (or two, if you increase flow rate) will affect water temperature everywhere. So, no.

The fundamentals of heat transfer tell us that the water exiting the waterblock is now at a lower temperature. Since we are now starting with a lower temperature for the radiator, the performance is affected by increased flow, *and* decreased temperature.

You're pretending two mutually dependent events are independent. They are not. For constant heat rate, the heat transfer coefficient and the delta-T are inversely proportional. Delta-T (and thus water temperature) decreases *because* the heat transfer coefficient increases with increasing flow rate. They are cause and effect, not competing causes.

Since a temperature decrease will reduce performance, the benefit from increased flow will be countered. Yes, the radiator can now move heat faster, but the waterblock's performance increase is still bigger. The radiator will get more heat than its performance increase can account for. The bottom line is that the radiator is going to have to work harder.

You're saying that increasing flow rate increases the power output of your CPU. Think about that for a moment.

Furthermore, the decrease in radiator temperature means that the radiator/fan relationship is not quite as good as it used to be. This fact exacts another toll on the radiator. The result from these additional influences is that the radiator is walloped with lots of extra work.

If you do nothing to the airflow, then the temperature difference between the radiator tubes and the air will not change. Altering the water flowrate only changes the temperature between the water and the radiator tubes.

However, if the radiator can't rise to the challenge and can't remove the heat, then it will leave the heat in the water. This will undo the flow benefit, and temperatures will rise to the point where there is once again balance between the heat taken from the CPU and the heat expelled by the radiator. This balance will occur at a higher CPU temperature.

Increasing flowrate increases heat transfer. Increasing flowrate causes adverse effects to cooling *only* when you are putting in enough energy by pumping that the improvement in heat transfer coupled with the increased heat load yields higher temperatures.

Alchemy