View Single Post
Unread 06-24-2007, 10:10 PM   #11
Cathar's Avatar
Join Date: Sep 2002
Location: Melbourne, Australia
Posts: 2,538
Default Re: Koolance's "how to test a radiator" demonstration...

Originally Posted by bobkoure
IMHO, this is exactly the kind of thing that might work one way under automotive conditions and another way under PC cooling conditions.
You've pretty much hit the nail on the head here.

The brass tube -> solder join -> copper fin, interface becomes more of an issue at very high air-flow speeds, such as those experienced in vehicles travelling at 60-80mph. At PC cooling air-speeds, where the low speed of the air-flow is by far the predominant factor in the thermal performance of the radiator, the impact of the tub->fin interaction is immensely reduced. Given common characteristics and materials used, in typical PC fan power applications, I calculated that at worst it accounts around 1% of the total thermal resistance of the radiator. Using 200cfm fans, maybe around 3%. As fan speed/air-flow increases, the problem gets bigger. For PC's, >100cfm 12cm fans are pretty much intolerable.

As for Koolance's marketing. After factoring in all of their stated test parameters, it turns out that for various data points in their results, for that to occur, the average radiator discharge air-temperature would need to be 15-20C hotter than the water intake temperature. Further, none of their liquid pressure drops at 5lpm even remotely co-incide with values independently established by testers. We're talking 10x liquid pressure drops to what has been established. Air-flow pressure drops are also dramatically different to what would be expected. It's as if there are major systemic errors in their testbed that somehow are variable from test to test. The more cynical person would go so far as to suggest that the either the competitor testbeds were deliberately sabotaged, or the Koolance radiator specific tests were run with a MUCH higher air-flow than stated.
Cathar is offline   Reply With Quote