PiSan wrote: wrote:
msmrodan wrote: wrote:
perhaps the relatively low waterflow actually *helps* the cooling since the hot water coming from the cpu has time to circulate and rise inside the reservoir instead of being sucked right back into the pump..
Actually that's not true. There is a greater thermal transfer when there is a greater temperature difference. For instance 10°C water flowing over a 40°C CPU will absorb heat faster than 20°C water.
Therefore if you have a faster flowrate, you'll be keeping the temperature difference high and increasing the rate of thermal transfer.
I'm not sure thats the full picture. I would agree that the thermal transfer rate is greater when there is a bigger temperature difference. The problem, as I see it, is which thermal transfer interface is the most important? You will get better radiator performance with hotter water, but lower CPU performance.
If you increase the flow of water to the CPU block then the water temperature leading to the radiator will fall as there is now more water to dissapate the same amount of heat. This lower temperature will cause the radiator to not function as well, leading to higher water temperature flowing from the radiator to the CPU.
Somewhere along the line there will be a happy medium, but I've long since lost the ability to calculate where this optimum flow rate will be. Maybe someone who is better than me at technical stuff would like to go into more detail about what the likely optimum flow rate would be?
If you had infinite water flow through the system then the water temperature would be constant throughout. Would this be better than a lower flow rate that had hotter water going into the radiator and then cooler water flowing to the CPU?
Hmmmm, this is making my brain hurt - I think I'll stop now........