While I have done radiator shroud testing before using more conventional 7 bladed fans, I wanted to do an update to both serve in shroud results and also give me some hands on time with my new radiator testing bench to see if there are any tweaks remaining before settling on a test method. I had noticed in my previous shroud testing that push vs pull optimization was different depending on the fan used, but I didn’t really have enough data to conclude it was fan speed alone or if it was simply the type of fan.
After testing many more fans, the Gentle Typhoons were again one of the stronger in CFM/dbA ratio on a radiator, so I figure it only makes sense to follow up with that fan to see what if any change is apparent there. In addition I used TFC 30mm shrouds in the old test and I wanted to compare that shroud to using an old gutted 25mm fan which is cheaper and many people do use.
So with that, I’m updating to a new radiator test bench with more controlled air flow measurements and the ability to also measure air flow volume.
Before getting started, I’d like to thank the sponsors that provided parts I used in this test:
Swiftech sponsored the MCR120QP radiator
Koolance sponsored the PMP-450 pump
Danger Den sponsored the MC-TDX waterblock
TFC sponsored the TFC 30mm shroud
TEST METHOD
Using my new radiator test bench V2, I’ll be running the system with the pump at setting 3 which is producing approximately 1.7GPM or so. Heat is provide via a modified aquarium heater and tuned to 125W for approximately a 10C delta system with the fan running at full speed. Seventeen thermal probes are measuring water in, water out, air in, and air out. In addition a hot wire anemometer is measuring air velocity, a “Kill-a-Watt” is measuring heater wattage, and another multimeter is measuring voltage at the fan plug.
Here is the setup with the access window temporarily opened. During testing the top acrylic panel will be closed so all air out goes through the air out port for air measurement.
I’m looking to log temps for about an hour after stability is reached to average out the ambient fluctuations.
As with any good testing, I think it’s important to show not only the testing rig, but also the data collected to develop the conclusion. While many go straight to the results, I’m sticking to my usual “testing story” approach and going to share the details of how I came to any conclusions.
