Cool It
Words & Images: Norm Sanders
When you pour all that petrol (and dollars) into the tank and start the engine, you are getting only about 30% value for your money. About one third of the energy generated by combustion is converted into useful rotational energy at the prop shaft. The remainder is lost due to waste heat, friction, and engine accessories.
All this heat has to be disposed of somehow, or the engine will literally burn up. Since most of the heat is created in the combustion process, the cylinders and heads are the hottest and need the most cooling. The needed cooling comes from the air, either directly or indirectly through water flowing through a radiator.
Direct air cooling has always been associated with low weight and simplicity since no radiators, hoses and coolants are necessary. Pioneer engine designers were well aware of this and one of the earliest successful air-cooled aviation engines was the V-8 Glenn Curtiss used to power the June Bug in 1908.
Unfortunately, the cylinders in those days were cast iron and the fins were very thick, due to limitations in the casting technology. Valve life was short. Mixtures were extremely rich so the fuel itself could help with cooling.
Just before World War I, the clever French came up with the idea of bolting the prop to the crankcase and rotating the engine itself to get more air past the cylinders. The rotary engines were far more reliable, often going 10 hours before overhaul. Spinning the engine worked for cooling, but had huge (and often fatal) torque problems. In addition, rotaries burned a lot of oil which was mixed in the fuel like a two stroke. The castor oil was less combustible than the fuel and was spewed out of the exhaust into the slipstream where it coated the aircraft (and pilot). As one writer delicately observed at the time, ”Castor oil is known for its purgative qualities. It would be impossible to expose oneself to such an atmosphere and not experience certain difficulties”. Many other early engine designers worked on liquid cooling. The obvious advantages were a decreased chance of shock cooling an engine, the ability to direct dedicated coolant flow to critical areas in the cylinder head such as the exhaust valve seat and guide area, flexibility in radiator placement, greater structural rigidity in the engine, and having the option of designing airframes with a relatively small cross-sectional area which could still house a powerful engine. There is, as they say, no such thing as a free lunch. Early liquid-cooled engines had added weight, more chance of battle damage in military applications and greater system complexity.
At first glance there still seems to be the choice between air and water cooling. However, the Rotax 912 series engines combine the advantages of both types. The 912 has air-cooled cylinders with liquid-cooled heads. This minimises the danger of burned exhaust valves, a bane to all aircraft owners. Liquid cooled heads are also available for Jabirus.
Continental and Lycoming have spent millions of dollars solving their valve problems. Those of us with Jabirus and VW conversions have to do our own experimentation.
My Sonex
I imported a Sonex from the U.S. several years ago. It has a Casler/Hummel 2.4 engine. (The plane itself was built by Scott Casler and Morry Hummel.) I was struggling with high cylinder head temperatures and burned an exhaust valve in the first five hours. Apparently Australia is a bit hotter than Wisconsin.
Obviously, my CHT was way too high at 420F. (215C.) Some VW sources say this is permissible for less than five minutes. However, 420F is valve-burning territory and is getting perilously close to the temperature where aluminium starts to soften, warping heads and letting valve seats get hammered down. I won’t go there again. Jabiru uses 392F (200C.) as a maximum and that became my target.
First thing I did was to get a plenum kit from Aerovee for $100 plus freight. This is a collection of baffles surrounding the top of the engine to channel air between the cylinders, just like Lycomings. I sealed all the openings with silicon and fired up the engine: Still running CHTs over 400F on climb.
Back to the drawing board. Maybe the air was bouncing off the back wall and not flowing past the rear of the back cylinder. I made up vanes in the corners to channel the flow. No improvement. Then I noticed most GA planes had a major portion of the front cylinders covered so more air would flow to the rear. I installed the baffles on the front. After that, I started thinking about where all that air was going.
Next step was to make up some permanent cowl flaps and cut out enough lower cowling to equal approximately two times the area of the inlet holes, because the air is moving more slowly after squeezing past the cylinders. While I was at
it, I cut inlet holes in the cowling just in front of the valve covers. VW experts say to throw away those nice, shiny after market covers and put the black ones back on so the covers can radiate heat (The standard covers are better at stopping
oil leaks, too). Of course, in a tightly cowled engine, there needs to be airflow over the covers. With sabre saw still in hand, I also cut vents in the cowling forward of the ignition coils to keep them from overheating.
Temperatures were better, but still too high. About this time, we were having the same problems with CHTs on our Jabiru-powered motorglider. The solution was to thin out the full-power portion of the needle in the Bing carburettor so that the mixture was very rich at wide open throttle. There was also a diffficulty with unequal mixtures (and temps) in the cylinders, which was solved by putting a cruciform vane in the intake manifold behind the carburettor to stop the vortices.
Running a rich mixture to cool the valves dates back to the early days of aviation and continues to this day in War Bird type recips. The air-cooled bombers, fighters and transports all leave trails of black smoke on takeoff with their mixture controls set at full rich.
The Sonex has an Aerocarb which turned out to be easier to deal with than a Bing. I adjusted the needle to a very rich mixture. This allowed me to use the mixture control to yield the desired temperatures by leaning from over-rich. Voila! Maximum temp. was now the magic 392F (200C.) and cruise well below Jabiru’s recommended 356F (180C).
The engine has run for 50 hours since burning the valve. Compression is good and valve clearances are stable. I set them at .008in (0.203 mm) which is slightly wider than normal. This aids in valve cooling because the valve is in contact with the seat for a longer interval.
I burn 98 octane Mogas in the Sonex after a bad experience in another VW-powered aircraft with Avgas. My CHTs were fine, but I burned an exhaust valve. I rang Great Plains Aircraft in the U.S for advice. “Are you burning Avgas?” he asked in an accusing voice. “Yes”, I replied meekly. "Well don’t. Avgas has far more lead than a VW valve is designed to take. It builds up between the valve and the seat and the valve eventually burns.” Yes, it does.
So now, after a lot of hacking and cutting and arriving at the right (rich) mixture, I have defeated the 70 percent heat monster. In all my struggles to deal with waste heat, I have revisited many of the efforts of the early engine pioneers. I didn’t however, venture into castor oil territory.