The Avid Flyer Mark IV cowl was originally designed for the Rotax 582 engine. It has two large cooling air intakes (two 240 sq cm / two 37 sq in) at the top of the cowl. With a 582, a water radiator is mounted behind each intake. There is no oil cooler.

A 912 requires air, water and oil cooling. On this particular Avid Heavy Hauler, a separate lower air intake (measuring 250 sq cm / 39 sq in) had been added under the prop. Both the water radiator and oil cooler were mounted one behind the other (water radiator in front) at the lower air intake, resulting in oil temperatures that were too cold. In addition, the upper air intakes were providing some air cooling of the cylinders, but were also directing a large volume of air to the carburetors that resulted in carb icing in cold, humid conditions, an unusual occurrence with a 912. A temperature probe installed near the carburetors confirmed that the air reaching the carb intakes was not being warmed above OAT.

At first, the oil cooler was moved in front of the water radiator, where it would be easier to partially block it in cold weather conditions and thus raise the oil temperature. However, any attempt to block the oil cooler also blocked the water radiator, resulting in coolant boiling in climb.

Then the oil cooler was moved to the front of the firewall, in the path of warmer air exiting the cowl. This improved the oil temperature in cold weather and eliminated interference with the water radiator, permitting independent blocking for temperature control.

In winter, the water/CHT temperatures were still too low at cruise ( both around 60°C / 140°F for OAT of 0-10°C / 32-50°F) and getting VERY cold on descent. It was therefore difficult to manage CHT : getting a good CHT in climb resulted in too cold CHT in cruise; blocking the water radiator to increase CHT in cruise created overheating in climb, to the point that the coolant boiled and completely filled the overfill bottle.

To correct the carburetor icing tendency and improve airflow over the water radiator at the lower air intake, the upper air intakes were blocked. Four round 50 mm / 2 in intakes were created, one for each cylinder. Scat hoses direct ram air to the cylinders (for cooling) where it is warmed before it reaches the carburetors. The remaining air entering at the lower intake is warmed as it passes through the water radiator.

The 912 is now running much smoother with warmer air at the carburetors (10°C / 20°F above OAT). Water, CHT and oil temperatures are now more stable and within recommended ranges. Coolant boiling is no longer an issue. Before these modifications, the area of the cowl exit was smaller than the total area of the upper and lower intakes, which restricted the airflow into the cowl. This restricted airflow was distributed proportionally among the large upper air intakes and the lower air intake. Most the the incoming air flowed towards the carburetors, less flowed over the water radiator. After these mods, the water radiator gets more airflow and thus the coolant is not overheating in climb.

During a six-hour flight to 12,000 ft (OAT 10°C / 50°F on the ground, -8°C / 15°F at altitude) climb CHT was 95°C / 200°F, oil temperature was 105°C / 220°F; in cruise, CHT was 70-85°C / 158-185°F, oil temperature was 90°C / 195°F. For winter operations, 10-20% of the water radiator and 75% of the oil cooler must be blocked.

The upper inlet area is now 80 sq cm / 12.4 sq in (4 holes of 5 cm / 2 in diameter) and I think could possibly be reduced further for winter operations. The Rotax 912 installation manual calls for a cylinder cooling section of 100 sq cm / 15.5 sq in, yet the ram air scoop that Rotax offers as an option for this engine (similar to the Jabiru ram air duct) has an inlet area of 50 sq cm / 7.8 sq in, that's half of what the manual recommends. Go figure out...

Further development will include a cockpit controlled cowl flap at the cowl exit to restrict airflow for warmer temperatures on descent. Carb heating will be installed in the near future as well.



Sept 1st, 2009 : more than 30hrs have been flown during the summer months. OAT temps encountered were mostly around or exceeding 30°C/86F. Engine cooling remained very satisfactory with the mods described (obviously restrictions on oil and water radiators were removed):

on climb: CHT=90°C/195F, oil=120°C/250F (max oil temp allowed by Rotax is 140°C/285F)

in cruise: CHT=80°C/175F, oil=100°C/215F

During this period I monitored closely the temperature under the cowl (I have a probe installed by the air filter).

The temperature was between 45 and 55°C (115-130F) which seem quite a lot, but I did not feel any noticeable power loss in takeoff and climb. I wonder if there is any official Rotax graph/study of engine performance vs air temperature.

I am considering creating two airscoops that could be opened in flight to blow fresh outside air onto the air filters for takeoff and climb, but I am not persuaded they will change a lot in performance.