Calculating the ( rule of thumb) radiator size of approx 600 cu" for 200hp, giving 3 cu" per HP, the size of the Mazda oil cooler then gives a .8 per cu" per hp. I wondering if this holds true for the 20B and indeed the single rotor.

George ( down under) 

Just an update on my RV-8 / 20B  oil cooling experiments. 

On the theory that airflow patterns inside the cowl were blocking airflow through oil cooler, I installed a partial exit duct behind the radiator directing the airflow downward toward the cowl outlet.  It looked very restrictive but flight tests showed almost no affect on water cooling (which is OK)  but a significant improvement in oil cooling.   I further restricted the airflow through the rad by putting some roof ridge vent material inside the inlet diffuser.  This gave a tiny increase in water temp but a further improvement in oil cooling.   Long story short,  after several more tests it became apparent that back pressure under the cowl was having a major effect on the oil cooling.   I have no idea why my instrument did not read the pressure correctly.  It works fine on the bench and is properly referenced to the static system in the plane.   The temptation is to keep changing the cooling outlet scheme until the internal cowl back pressure is low enough to get the cooling good enough.  My belief is that this would lead to a very high drag solution.  You may remember the experiment I did by flying with the cowl removed.  The cooling was never a problem then (except perhaps too much cooling) but the drag was enormous.  The fuel burn was 60% higher at the test airspeed of 130 mph.

The conclusion I eventually came to was that the rad (because of it's relatively low air flow resistance) is hogging the airflow capability of the cowl cooling outlet.  (cowl flap did not have enough effect to fix the problem).   Keep in mind that the oil cooler is a thick AC evaporator core that is very restrictive.   The current experiment is to replace it with a much less restrictive (to airflow) oil cooler.  I found the largest cooler that would fit in the same location as the AC core and I'm using the same diffuser as before (slightly modified to fit the larger face of the new cooler).  This cooler is only 2" thick and core volume is 30% less than the AC core.  It is slightly larger in volume than an RX-7 cooler.  Without any back pressure (flying with cowl off), the AC core had way more than enough cooling capacity (146 F oil temp on a 93 degree day) so I'm hoping that this smaller cooler will be enough.  Should be ready to flight test it this week.

I should point out another symptom. Power setting (and therefore airspeed) had very little effect on the cooling  (i.e., it didn't get much hotter at high power as long as airspeed went up as well.   Things got hot fast in climb however.  This also indicated to me that cooling was limited by airflow through the system rather than by the oil cooler's ability to transfer the heat to the air.  If the cooler is simply too small, more airflow will not help much. 

Tracy