There is something to be said for this way of mounting radiators.
Back in World War One almost all liquid cooled aircraft engines
had flush mounted rads because they worked.

The main rule is none of the rad area can be blocked. This installation
violates that rule. On the up side, the duct is near 100% efficient
with near total recovery of dynamic pressure. This means extensive
external diffusion is taking place depending on the thickness and restriction
of the rad to airflow passing through it. The other rule is of course the 
restriction to airflow out of the cowl must be at a minimum when maximum
cooling is required. Cowl flaps will reduce the drag and cause more external
diffusion when max cooling is not required.

The down side is the area required is too large and hence the drag is
too high when used as the sole source of engine cooling. In this case only 
the cyclinder heads of a Rotax 912 are being cooled and I am not sure how well.
 
Paul Lamar
 
The AirCraft Rotary Engine NewsLetter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org