Impact of flow in a radiator, any review?

[prava]

Hi folks.

As of now we have plenty of information about nearly anything...but I can't find any serious review about how important is flow for the performance of a radiator. Do you know of any review about it?

Cheers :up:

[Nickel020]

I think Martin did something like that or mentioned it in one of his rad reviews, look it up. Basically he found that unless flow is ridiculously low (like less than 0.05GPM or so) there's absolutely no difference.

[mrcheeky]

+1 to above answer, except I remember it being something like 0,5-1GPM after it makes no difference to radiators performance.

[relttem]

in simplistic terms, it is just a energy balance

h(air)*Area(outside radiator)*(Tair(out)-Tair(in)) = h(water)*Area(water side)*(Tin(water)-Tout(water))

as flow slows, h lowers, so you have to balance it with a change in Touts since A and the Tins are constant (but the water Tin will change too b/c it is now at a higher temp going back to the CPU). If you slow it enough you get a runaway system

As mentioned above, it might not have much of an impact at all.

[Vapor]

Martin showed (on just one radiator--each radiator model has a unique flow vs. dissipation curve) that above .5GPM, the scaling is flat and extremely minimal (going from .5GPM to 3.5GPM improved dissipation a mere 5%). The curve also depends on airflow but, generally, it doesn't even 'curve' until extremely low flowrates.

http://img707.imageshack.us/img707/3298/badiz.png

[skinnee]

I am changing the Skinnee Labs radiator test procedure for variable flow rates, but I have at least 1 (probably two) more radiators to test before I change the procedure and no I will probably not go back and retest all the radiators already reviewed as I do value my sanity. The main reason is for another project we're working on, but its good data to have and also verifies data reliability in testing.

[prava]

Thanks folks. So paralel rads seems quite an interesting idea as you lower restriction of the overall loop while maitaining performance on the rads themselves...