General

The EG33, unlike the Lycoming and Continental engines, is a liquid cooled engine and as such requires auxiliary cooling mechanisms in the form of radiators, water pumps and a few other elements to make an effective system. The automotive industry has mastered this concept over the past few decades and the technology has moved forward in leaps and bounds. Much of this technology is transferable to the aviation application. The biggest problem that faces the aircraft application is to produce an effective cooling system that does not add too much weight and drag, both of which are detrimental to the performance goals of the aircraft.

Design Issues

The standard air cooled aviation engine is designed to operate at a temperature significantly higher than that of the liquid cooled counterpart. This elevated operating temperature has the detrimental effect of reducing engine life, consuming more fuel, and being susceptible to shock cooling, however the higher difference in temperature between the cylinder heads and the cooling air makes it a more effective heat transferring device. The liquid cooled engine, because of it's lower operating temperature, transfers it's heat less effectively, to the environment however with suitable radiator design the cooling performance may surpass that of the air cooled engines thereby extend the engine life and reduce pilot workload.

If one considers an aircraft traveling at 200 mph with an intake area of one square foot, then, assuming that the cooling air passes through the intake unimpeded by the cooling arrangement, 10kW (13.7 hp) of heat is transferred from the engine to the environment for every degree Celsius temperature difference between the input and output temperature. Thus, if the ambient temperature is 15 degrees Celsius and the cooling exhaust temperature is 100 degrees Celsius (Dt = 85⁰C) the most heat that can be extracted is 868 kW (1165 hp). The real world situation is that cooling air is impeded and the exiting air is not at the same temperature as the radiator transferring the heat.

How much heat has to transferred to the cooling air is somewhat of an unknown. A gasoline powered engine has an efficiency of approximately 30% and of the 70% inefficiency 35% is said to exit directly from the exhaust pipes in the form of heat, kinetic energy, and sound leaving 35% as heat in one form or another. This heat is further subdivided into the liquid cooling afforded by the coolant, liquid cooling through the oil and subsequently transferred to the oil cooler, and direct cooling through the cylinder block. (Aluminium engine blocks have radically improved the cooling efficiencies of automotive engines over the past two decades.) The cooling system must therefore be capable of removing as much power in the form of heat as there is turning the propeller. If we make the following assumptions (Typical Cozy dimensions):

1. The air is unimpeded

2. Cooling exit temperature 70⁰ C

3. The intake has a width of 15 inches, and

4. The intake has a height of 6 inches

then the following table illustrates the power/cooling requirements for three scenario's of climb, cruise, and max speed.

The last column indicates the excess cooling power available with the above three assumptions. Clearly the worst case scenario is in the climb where only 55% cooling excess exists. There are, however, some factors working with us during the climb. The adiabatic lapse rate of 2⁰ C per 1000 ft of altitude results in a further 11hp per minute increase in efficiency as the aircraft climbs, meaning that the cooling excess is 60% at 1000 ft of altitude. Likewise engine power, and consequently cooling requirement, also reduces with altitude so there is a similar impact resulting from altitude dependent power loss however the loss in air density on cooling efficiency offsets this latter property and thus the power/altitude argument must be disregarded.

The goal is therefore, to design a cooling system that does not consume more than the 55% cooling excess that is offered by the flight characteristics and the cooling aperture.

The Design

The above analysis indicates that there is no fundamental reason why the engine will not cool adequately given that sufficient airflow and heat exchange is possible. In the climb condition a four inch tall aperture would be on the upper theoretical limit for cooling an engine of this power output however an air cooled engine could potentially use a smaller aperture given that the glass fibre cowling could stand the high temperatures of the cooling exhaust.

Efficient cooling requires, not only efficient air entry but also good scavenging of the cooling air from the cowling plenum. This scavenging must also be set up for a number of conditions, in particular ground cooling where there is no airspeed, and during the climb where the cooling excess is marginal. Conversely at high speed a large intake area induces considerable cooling/airframe drag and is to be avoided.

The following design adopts three primary techniques to optimise the performance under all conditions:

1. Variable and adaptive intake geometry,

2. Trailing edge cooling exhausts,

3. Lower cowling exit, and

4. Exhaust augmented cooling (courtesy Charlie Airseman).

Variable and Adaptive Intake Geometry

The standard Cozy MKIV has a belly mounted NACA scoop which is 15 inches wide and about 4.5 inches in height at the throat. Some builders have made this smaller and other builders have made it slightly larger but 4.5 inches is a good average for the vertical component of the opening. As we have seen in the preceding discussion the aperture is marginal during the climb but more than adequate in the cruise condition. Numerous builders have sited cooling issued in protracted climbs particularly in the warmer latitudes. Typically the overheating, during the climb is compensated by reducing the rate of climb or backing off the power, however the other end of the spectrum has not been addressed where excessive cooling drag is encountered in the cruise condition. To overcome this paradox I have constructed a variable geometry aperture that is coupled to the engine coolant temperature monitor to provide temperature regulation. Essentially the design utilises a false bottom to the NACA scoop which is hinged, using a fibre-glass spring hinge, at the front of the scoop. The rear end of the scoop is moved, in the vertical direction, using a motor drive such that a full 6 inches of opening may be attained for maximum cooling, and total shut off for rapid warm-up in Canadian winters. To improve the effectiveness of the scoop the sidewalls have been extended below the fuselage bottom this can be likened to adding VG's in front of the scoop as so many builders are presently doing.

The picture to the right illustrates the belly of the beast. The raised sidewalls of the NACA scoop are clearly visible. The front most part of the scoop, the sidewalls, are at approximately 60⁰ w.r.t. the horizontal. This angle is maintained over approximately 80% of the length of the scoop where after the angle increases to almost vertical. The intent is to make the sides of the duct/scoop act like a vortex generator thereby inducing the high energy air into the duct. At the back of the duct there is little VG effect and the intent is to deflect the air into the two ports (one either side of the duct) which are used to provide engine air and oil cooler air. The bottom (top in this picture) of the scoop/duct is a horizontal aerofoil section to minimise drag. The lighter area of fibreglass at the first 6 inches of the duct is reinforcement and imbedded nut-plates to accept the attachment of the moveable NACA floor.

The motor drive is shown, centrally located behind the rear fixed gear bulkhead (n.b. the space between the two fixed gear bulkheads is used as a sump tank into which the two primary strake tanks, discharge prior to being pumped to the engine.) An aluminium ring machined from 2024 aluminium distributes the load for the three screws that attach the motor. The motor is from an automotive electric window winding mechanism. The running current is 2 Amps and will stall at about 15 to 20 amps. A 5 Amp polyfuse protects the motor in the event of jamming.

This photograph shows the NACA moveable bottom in it fully up condition. The lead screw can be seen in the centre of the duct. There ate two thrust bearings in the motor mechanism so no additional thrust bearings are needed in the lower bearing located in the aerofoil section. The lead screw uses a 7/16" UNF thread.

The distortion of the NACA sidewalls is an illusion and caused by curvature in two planes. When seen by the naked eye the shape is very apparent.

This photograph shows the duct in the fully closed condition. This condition will not happen in practice and there will be a limit switch to ensure that some air is always available.

The parts necessary to make this mechanism are shown in the adjacent photograph and are all capable of being made on a hobbyist's lathe.

The system becomes adaptive through the use of an electronic controller which positions the vent according to the coolant temperature. The standard coolant thermostat is removed so that the engine tends to run cold and is offset by a reduction in intake aperture. This minimises drag for most conditions except for protracted climbs in high temperature environments. Under most conditions the cooling drag is significantly reduced.

Trailing Edge Cooling Exhausts

One of the greatest problems with the cowling exits, found on the EZ range of aircraft, is that the propeller is directly behind the opening. The propeller produces a bow wave in front of each blade which pushed air back into the aperture. On average air exits through this opening but with so much "to-ing" and "fro-ing" efficiency is lost. Many builders have done oil drop tests to demonstrate this problem. The cooling efficiency is dependent on the cooling exhaust as much as the intake design. In an attempt to remove the bow wave problem and make an efficient cooling air exit, some trailing edge exit vents have been fabricated and can be seen in the lower right hand side of this image. In this design the air flowing over the upper and lower surfaces, of the cowing part of the strake, cause air to be sucked out from the cavity behind. This cavity is connected to the main engine bay and results in a general reduction in pressure within the cowling. Because of the design of the lower cowling and the radiator arrangement. This causes air to be principally drawn through the upper radiator. By placing the trailing edge vents 21" in front of the propeller the air should be sufficiently straight to avoid upsetting the propeller performance.

It is the intention of this design, to provide suction during ground operations so long as full power is not applied. Hopefully this design should provide sufficient cooling during run up and taxi operations.

Lower cowling exit

The lower cowling exit exhausts air predominantly from the lower radiator and is sufficiently far forward from the propeller to avoid the bow wave issue described above. The large 12" hole from which the propeller shaft extends is unfinished at present and will be almost entirely blocked and a matching spinner will be added to ensure effective streamlining properties to the tail end of the fuselage.

Exhaust Augmented Cooling

I first heard of exhaust augmentation cooling from Charlie Airesman who had successfully applied this technique to his Vari-eze. Essentially the exhaust is inserted in a throat of a large tube and pumps the air from the plenum behind. In the picture above the tip of the exhaust pipe is visible. This is not in its final state and requires about six inches of exhaust to be removed for effective operation. This method of air extraction is particularly effective since the maximum effect is achieved at full power when maximum heat removal is required.

Radiators

The radiators are attached to the lower engine mount in a wedge format. The radiator cores come from a VW Sirocco and were found new at a cost of $85(Canadian) each. There are various sizes available but these have core dimensions of 21" x 13" x 1.5" (length x width x thickness) The total cooling area is therefore 3.8 square feet and the volume is 0.474 cu ft. Typically a cooling area of one square foot per 100 hp is the magic number for cooling area so this design should be good for more than 350 hp or conversely adequate cooling should be possible even in the hottest climates.

The VW Sirocco radiator was selected because of its cross-flow design which gains approximately 15% over a top down design. The following illustration shows the difference between the cross-flow and top down radiator designs.

Conclusion

The cooling discussion given above has yet to be verified but is founded on a number of data points of fellow builders. The design provides cooling for both ground running as well as cruise and climb conditions. Hopefully the adaptive intake geometry and the ultra clean termination of the streamlines over the propeller, should yield a very low drag result.

Last Updated:    Thursday August 31, 2006