Login Form

Q-talk 93 - My Cool Onan

Dennis Clark, Newnan GA

As I was building my Quickie, reports of hot running Onans were numerous. The early reports seemed to blame the original 18 HP heads, so I dutifully bought the newer heads. Others, even with the new heads, still reported overheating. Revised baffling was the next suggested fix for this issue, however, it didn't help, according to later newsletters. Builders were going through mental gymnastics trying to solve the problem, even envisioning a welded-on water jacket (envisioned, but not done).

From the age of 12 or so, I was very interested motorcycles. Back then there were no water-cooled motorcycles. All had the exhaust facing forward to cool it first, except the rear cylinder on Harleys. Drawing upon this experience in the attempt to solve the Onan overheating problem, I even "envisioned" the Onan spun around 180 degrees to get the hot side (exhaust) forward. To do this, however, too many mechanical revisions would have to be made.

Then I had this idea, route the air to the backside first. The more I thought about it, the better it seemed. I could think of no negatives, except more fiberglass work. As my scheme progressed, it appeared that I would have to duct the air under the cylinders and around the oil sump. While I was at it, I figured I could blast the sump with that cool air to lower oil temps.

Now to fill the duct with air, I needed an opening. The simplest thing was to converge the two ducts in front of the sump and cut an opening under the spinner. In doing some research, I found that A/C pressure point representations indicated high pressure below the spinner. Good! Now it was time to get to work!

I chose the "smile" inlet because it was low on the cowl where I needed it and I could use the toothy mouth you see painted on some WW2 fighters. If I had used the widely spaced round inlets (LoPresti), I might have had two extra 90 degree turns added to the ductwork on each side. These ducts do eat up interior room and with our narrow cowls, we might not get the same benefit as one for a wide VW installation.

Anyway, my theory was to hit the hottest side of the engine with the coldest air I could get. Onan must have had similar thoughts. With the QAC set up, you can have cold air flowing down the back (flywheel side), but air flowing through the cyl head fins is heated by passing over the intake side and then the exhaust. The intake side is cooled by the intake charge. It needs very little cooling air. Air passing through the intake fins picks up heat and by the time it gets to the exhaust side, its ability to remove heat is limited.

The nose shot shows the interior of the duct. Cooling holes are cut for the oil sump. These were enlarged at the airport with hand tools so they are a bit sloppy. They are about 3 sq. inches in area. Two more are cut after the duct makes the bend around the sump. These are about 1 sq inch each. I did no testing to optimize these openings, but I am happy with the oil temps. Except for an outlet to the carb, there

are no other escape paths for the I air. The carb heat box has a "V" shaped flapper that closes the hot air path when normal air is selected. This allows whatever ram pressure that is generated in | the air inlet to be used by carb.

The cowling has a second lip around the inside of the mouth to create a capture point for the baffle seal.

The duct is split left/right just below the prop shaft, with a flange overlapping the opposite side to attach the pieces together. Nutplates are handy here.

The feet are removed from the sump to give a smoother airflow inside the duct and save an ounce or two. The outer wall of the sump at the sump mounting bolts remains to provide an attach point for the duct. A nutplate is riveted inside the two forward walls. The lower edge of the heads has an area that can be drilled and tapped for other mounting points. These can be accessed though the mouth of the duct.

As the duct enters the bend at the rear of the engine, I started decreasing the area of the duct. The width stays the

same but the depth is decreased. Working with the left cylinder first (looking forward), I smoothly blended it to conform to the shape of the Kevlar plate. The back of my duct has a diagonal step to limit the reduction in area at this point. The face of the duct that hugs the cylinder is terminated at the cylinder casting marks (about half way up the cylinder). You might worry about making the air reverse directions at this point but it seems to be OK on my engine. I considered buying a 10 mm CHT sensor for each stud just to see the temp at each area.

My spinner is 10" in diameter. The smile inlet was created by lowering the center of a 10" circle 3". This is a relatively large area. If I was to experiment with drag reduction, this might be the place I would start, while monitoring all temps on the head. This was my first attempt at revised cooling. It surprised me how well it worked. There may be room for improvement. Play with it if you wish, but try it as it is first.

The right cylinder duct is a mirror image of the left duct. This equalizes the airflow to the opposite cylinders and heads.

As you can see, the head baffling is terminated at the spark plug, I wanted rapid airflow in this area. Extending the baffling beyond the plug may inhibit flow. There is a step in the fins at the exhaust side of the head, on the same plane as the head gasket. I bent the baffle to not allow air to escape at this point. The large threaded hole at the top of the head is a logical attach point for the baffle. At the forward edge, I drilled a small hole in every second fin. Two corresponding holes in the baffle provide a way to secure the baffle to the head with safety wire.

The space between the duct and the cylinder/head baffling is covered with tape and glassed to provide an attachment between the two. Nutplates inside the duct and head baffling are used to secure the glassed attachment. A picture is provided. You will note that the attachment does not touch the cylinder fins but is very close.

A metal baffle under each cylinder runs from the casting mark at the center of the cyl front and rear. This to slow the flow in this area as it needs less cooling than the area at and around the exhaust. There is a lip at the forward edge of this baffle to conform with the shorter fins at the base of the cylinder. A seal of foam rubber is glued to the aft edge of this baffle to provide a seal between it and the glass duct.

At the top edge of the duct, baffle seal material is riveted to the duct to seal this area. With the engine inverted, I used a 3/8" sheet of foam to provide a mold for the bottom sump cover. The foam is used to space the glass away from the sump. Cover it with plastic to allow easy foam removal. Nut-plates pop riveted inside the duct secure the sump cover.

Temps? LESS than 300 degrees CHT at stud aft of exhaust valve and LESS than 200 degrees OIL

You can order a printed copy of Q-talk #93 by using the Q-talk Back Issue Order Page.