Engine Cooling Ideas

This blog covers the background behind a Rotax cooling system design

Overview
This builder would like see what the effect of closing up the front door would have on the Tucano. To do this requires a solid foundation as currently the gear opening is the primary means for ejecting the waste heat from the radiator, oil cooler, exhaust and other minor sources. It is hoped that this would remove a perceived source of drag as well as some cooling drag - maybe.

All the reading says that air outlet's have to be twice the size of the radiator area based on heating the air. This did not have a convincing ring and seemed to be a easy way of solving the problem.

John Thorpe stated in one of his many articles on his design philosophy for light aircraft  that the largest drag left in T18 was cooling drag but required careful design due to the danger of overheating. 

Minds much more qualified have discussed this issue in great detail but the short article below outlines the basis of the my current opinion.

Research
In roving I can across a article on the Merdith Effect used as the basis for the design on the P51 cooling system. The most interesting comment was that at full power at altitude the outlet was squeezed down to 5'' square - See article: Merdith Effect fact or fiction.

Explanation of the P51 duct operation

Position A: High speed air with ram pressure enters the diffuser
Position B: Air is slowing, pressure rising as the diffuser cross-section area increases.
Position C: Air is at minimum speed and highest pressure on entering the radiator. 
Position D: Air has been heated on passage through the radiator. Pressure falls a little due to viscous losses in the radiator. 
Position E: Thermodynamic expansion is taking place in the nozzle. Temperature of the air is falling as the heat energy content is converted to kinetic energy. Pressure is falling as the air speeds up down toward the exit.
Position F: Adjustable exit scoop is set for high speed flight. Temperature of the air is much reduced, but not all of the heat can be turned into kinetic energy. Pressure is close to ambient, while the exit velocity is close to matching the airspeed of the airplane. The momentum change going from radiator rear face to the exit is seen as thrust, cancelling the drag of the radiator.
Position G: The scoop is wide open for low speed operation, where the cooling needs must be met by low velocity of the air at position A. No thrust generated due to low pressure rise in the diffuser.

Beign forced to dabble in Bernoulli a long time ago it did have a ring of truth. More roving lead to a discussion at of all places a car blog that was saying the same things.  

The first diagram below shows what I would believe is happening with a lot of radiators installations. Air is like any fluids and dose not like work and will seek the path of least resistance so if nothing else a radiator hung on front of a engine with a big hole in front as its only protection it will be inefficient by the air looking for a easy way out - aka a big hole in the back of the cowl.

Extract:  "Then what you want is a Kuchemann & Weber trumpet shaped diffuser. The name comes from a couple of scientists, Kuchemann & Weber performed a lot of R & D on this subject during World War II and published their findings in Aerodynamics of Propulsion sometime in 1953. These diffusers are highly utilized in aviation as well as in automotive racing. 

The book is out of print and extremely difficult to find, except in college/university libraries. Anyway, their research showed that the following type of bell mouth intake is inefficient: 

 

...and that more of a trumpet shape prevents cavitation/turbulence and also traps more airflow, forcing it through the heat exchanger and preventing a lot of airflow from stacking up... 

 

Here are some CG images of a theoretically perfect Kuchemann & Weber diffuser: 

 

While you might not be able to achieve the exact shape that their calculations would generate, a rough approximation of using about 1/4 to 1/2 your radiator surface area (depending on your specific cooling needs, i.e. street or track) for the diffuser opening and then gently curving walls (like you're making a trumpet) that join to the edges of the radiator will get you a lot more efficiency than running without any duct work."

A trumpet is not practical but the diagrams would reinforce the idea that were outlined in the article on Meridths work. 

The final word belongs to Lee Attwood the designer of the P51 when he gave his last talk in England reported here in an article from Historic Racer.

Overview

What dose this this mean - with a cowl closed cooling air has to be ducted to achieve a acceptable level of cooling efficiency while minimizing losses and a large exit is not required in a well thought out and ducted system. Replica aircraft lose freedom in design so no great gains could be envisioned but if we dispose of the air in a controlled manner to cool the engine in climb it may be possible to close the door's on wheel well opening.

A water cooled motor has a singular advantage over a air cooled motor, you place the heat exchanger anywhere and with the Tucano that's is under the engine. This preclude's the use of a standard Rotax exhaust and requires two separate;exhaust and mufflers to free up the underside, such a system is installed on a Highlander located at Rylstone.

To concentrate the heat load a Silent Hektik water/oil heat exchanger would be installed so both these heat load's are at one location. Silent Hektik have stated that a single radiator will meet the requirements of a standard motor so a single radiator installation will have to be at 100% efficiency to cool a supercharged engine. 

As shown below the air is taken in, slowed to raise the pressure and extend the time for a volume of air to cross the radiator interface before being accelerated and dumped out of the two fake turbine outlets. 

Forget thrust as Merdith's work at the outlet side is only applicable to aircraft traveling in excess of 300 mph.

The outlets are fake turbine exhaust exiting though the underside of the fuselage.

The inlet and outlet are fiberglass moldings with the two 45 degree standard oil cooler ducts. Radiator is a stock Rotax with the connections 4'' scat hose. The outlets would have to be fabricated over two half CNC machined foam cores that should survive manufacture of two sets 

The duct assembly / radiator would have to be mounted independently to the engine using support rail's incorporated into the lower cowl supported by the firewall and rigid engine mount. The exhaust would be two independent systems to open up the base of the engine to allow the installation of the system.

Will it all fit - $64 question as there is a lack of 3D models so it will have to be checked by comparison to the installation on the aircraft. What I will have to do in the spare time when at home is a little more thinking with numbers. 

So the thinking is out there and I will await comments.