GlasairProject.com Subaru Cooling System Design

Cooling System Design (newest entries at bottom):

The cooling system has always been a source of huge debate among auto conversion "experts". My approach has been real simple from the beginning. The 200hp Lycoming has been shoe-horned into the Glasair for years now. None of these installations have needed any major cowling changes for adequate cooling.

So, I'm attempting to keep the cowling as bone stock as possible. I figure that the same amount of air that enters and exits the cowl and cools a 200hp engine, can cool my measly 165hp Subaru. It was all a matter of ducting the air through a radiator somewhere inside the cowl. The only side note to that is a small amount of cooling air to direct at the gear reduction box.

It took several years of thinking and measuring and re-thinking to come up with the cooling system that will be in my plane during the initial engine runs and taxi testing.

As for the radiator, I figured I needed at least 320 cubic inches of cooling area to adequately cool this engine. This was based on available radiator fin dimensions and spacing. I settled on a 3" thick core cut to a 5" x 24" size. I then had a double pass radiator fabricated by C&R Racing of Indianapolis.

The only location I could fit this radiator was up against the firewall inside the engine mount tubing. I designed a radiator shroud and the radiator to work together to place the radiator in this location.

Connecting the engine to the radiator was accomplished using an "L" bend piece of 1.5" radiator hose connected to an "L" bend 1.5" mandrel bent aluminum tubing with beaded ends. The other end of the tube connected to a 100deg piece of radiator hose that connected to the stock water pump inlet.

Up top was another story. I still haven't settled on a "final" design. What I have now works and it goes something like this; I retained the stock outlet crossover to keep the water exits balanced like the engine was designed. This crossover ends up pointing at about a 30deg angle towards the left side of the firewall. From the crossover, water flows into a capped swirl tank. Leaving the swirl tank it flows parallel to the radiator entrance, which was not the greatest idea. I thought it would be easy to find a 1.5" radiator hose with a 180deg bend in it. I was wrong, but, of course, I made due. I used another piece of mandrel bent aluminum tubing connected at both the swirl tank and radiator with short sections of straight radiator hose. Not pretty, but it works until I can fabricate a different swirl tank.

11/04/07: I fabricated a new swirl tank from the original tank. Now the outlet from the engine goes directly to the radiator inlet using two pre-formed radiator hoses and a short piece of aluminum tube.

The new swirl tank sits separately above the engine. A 1/4" hose from the high point of the crossover tube is connected to the front canister fitting. Another 1/4" hose comes from the high point of the radiator and is plumbed into the rear canister fitting. The bottom of the canister is plumbed with a 3/8" hose down to the thermostat housing to supply the thermostat with heated water for temperature regulation. I am much happier with this setup.

Ok, now I have the radiator, shroud, swirl tank, hoses and, OH, did I mention the in-cabin heater? This little gem is from Vintage Air. It has a three speed motor and molded plastic housing around the heater core and is quite small.

I mounted the heater on a pair of metal reinforced fiberglass channels molded into the fuselage sidewall. The hoses come through the firewall and are covered with fire sleeve both inside the cabin and the engine bay.

After all that, I now focused on getting the air into the radiator shroud. For this I decided to mold secondary walls inside the cowling, using the outside walls to direct the air back alongside the motor to the shroud. I fabricated transitions at both ends of the upper and lower cowlings to direct the air into these spaces and then to direct it from them into the shroud. This process took about 25 hours of fiberglass work to accomplish.

So far it's gone great during engine and taxi testing.

01/24/09: While the propeller and gear box were being repaired I fabricated a cowl flap. I utilized the pitch-change motor, from the damaged Quinti, as the cowl-flap control.

To measure the effectiveness of the new cowl flap I installed tubing with stone filters at the radiator inlet, radiator outlet and cowl inlet. I routed the tubes out through the oil filler door, along side the cowling and into the right side cabin vent. They are attached to a large Magnehilic differential pressure gauge for in-flight readings.

LATEST UPDATE 01/14/2010: This past summer we saw temps hit 110 deg f. in the Portland area. I did some long-climb testing and found that the engine would reach 225 deg f. during a long climb in these temps. This is when I decided to add more cooling capacity.

I began by taking a close look at my existing cowling for space. I really didn't want to modify the stock cowl anymore than I already had. Unfortunately there wasn't much room to accomodate this. I ended up compromising on a location under the gear box where I could get about 11" of duct ahead of the radiator if I fabricated a duct that would go through the lower cowl and forward to just behind the spinner line.

With the extra air coming into the cowl I needed to streamline the transition from the vertical firewall to the horizontal exit area. The first photo is a side view of the area showing the engine mount tube that will need to be incorporated into the transition part. She second photo is a top view. This is the direction the main radiator air is coming from as it tries to enter the exit area. Not very efficient at all.

This next two photos show how the new transition looks. I fabricated this by cutting a 3" diameter aluminum tub in half length-wise. I used a hammer and anvil to shape what would become the top of the transition. I increased the radius in this area so it will "reach" above the mount tube and transition to below the firewall. The extra piece of aluminum is to reverse the radius to direct the air away from the firewall and onto the new transition.

Next I focused on fabricating a duct. The duct had to have a small frontal area and expand to provide cooling air to the front of the auxillary radiator. I started cutting foam and hot-gluing it together to shape the inlet. Once I had the base of the inlet shaped I applied a two-layer laminate on the inside. I continued hot-gluing panels and laminated them as the assembly came together.

Fabrication continues with cutting the hole in the cowl and test-fitting the new duct. Once I was at this stage I hauled the entire assembly to the airport and trial fitted it to the plane to find the proper alignment to the radiator.

Fabrication wraps up with laminated the duct into the cowl then laminating over the outside of the cowl and duct.

And finally a sneak peak at the cowl on the plane.