Source: 2010 Q4 Beartracks, Kevin Deutscher
After a rather long build process I am finally getting to a point where I can no longer put off decisions about the fuel system and how to implement in my plans built Bearhawk. I am using an injected IO-470 and have to incorporate fuel returns, high pressure pumps, flow meters and a few other tidbits.
The really big decision is where to place the fuel selector and best locate the low point drains. I liked a panel mounted location, with a Left, Right, Both and Off positions. This seems to make sense and keeps the selector up and convenient both to view and operate with ease. I have seen several variations on this theme in several Bearhawks and the Stinson Voyager. They seem to work. My concern was how would the system perform with a fuel pump in the delivery line? To test out the system I mocked up the delivery lines per Bob’s Beartracks recommendations with 5/16 ID clear Tygon tubing and went about seeing how good the delivery flow was.When filling the lines from total empty the fuel would flow like the proverbial cow on a flat rock. I was able to start, stop and vary the flow rate at the fuel servo location at the upper back of the engine. So far so good, a solid baseline.
Inquiring minds wander and wonder and ……… what would happen if………. How about unporting a tank? After introducing air into the system, simulating an unport it was difficult to reestablish a solid fuel delivery until all the bubbles had been removed. The bubbles would tend to migrate back towards the tank but flow delivery was always erratic and feeble. Leaving the system to settle out for a while resulted in large air bubbles in the lines. Needless to say this observation begat several episodes of water play.
The results go something like this:
Single side fuel delivery was problematic due to air locking if unported. Both Sides ON was best at purging air and establishing unrestricted fuel flow. If the lines were unported single side it took a head pressure of almost ¾ of a full tank to establish flow and purge the lines. This would equate to 12 to 15 gallons of fuel in the aircraft.
Locating fuel selector up on the panel only aggravated the situation causing a large difficult bubble to form at the high point of the looped up system. Time for a little actual aircraft testing. Scott Williamson owner of N509RF had more than passing interest in this situation as 509 has a panel mounted fuel selector. And he is currently building a quick build kit. We devised a ground based testing scheme to see what would happen in the aircraft. Scott drained the left main tank leaving approximately 3 gallons and isolated it by placing the selector in the Left position, drawing fuel from the left tank. The engine started and operated normally during a long taxi out to the runway. Minimal fuel was visible in the sight gauge.
Approaching the runway we began a large radius sweeping right turn (taxiing on the ground) and sustained the turn for several revolutions. The centripetal force migrated the fuel away from the wing root, no visible fuel in the sight gauges, eventually it got quiet, the prop motionless. When the aircraft stopped moving, fuel was visible in the sight glass. Unport the tank and the engine will stop. Data Point!
Numerous attempts to start the engine resulted in ……… no start and a dead battery. Subsequent starting with the fuel selector returned to the BOTH position resulted in immediate starting. This test was repeated a few times with the same results. Unporting a tank in single side operation will create air bubbles in the fuel system. The fastest way to restore fuel delivery is to select BOTH on the fuel selector. This allows fuel to flow in from the highest fullest tank and vent bubbles up thru the other lines.
My curiosity has been satisfied, fuel selector on the floor and fuel line routing per Bob’s recommendations has the highest integrity based on bench testing and ground testing. Builders should follow Bob’s recommendations and utilize proper routing with smooth sweeps always flowing downward from the tank outlet, no S traps in the runs. Fuel selector should have a BOTH position.
1. The “Bob System” is optimal and should be kept as the “Gold Standard”.
2. Modifications to the fuel system should be builder tested to verify all functions.
3. The refill of the fuel lines after unporting a tank happens faster and more reliable when the fuel selector is in BOTH.
4. Take off and landing should always be on BOTH.
5. Sustained uncoordinated flight and or single side fuel delivery should be avoided while maneuvering in the takeoff or landing phase of flight.
6. Operation with minimal fuel levels should be approached cautiously.
A further extension of the fuel system thoughts might be to review what to do should there be a fuel delivery interruption.
Emergency Procedure – Engine Failure might look something like,
Wings Level
Pitch for best glide (75 mph)
Carb Heat ON
Fuel on BOTH ( Not fullest tank because that isolates the ability to move bubbles out.)
Fuel Pump On (If Equipped)
Mixture Rich
Prop Increase
Throttle Various
Ignition Check
Select Landing Site
I have resolved to install fuel selector on the floor, fuel delivery loops per Bob’s methods. Builders with a different scheme should test their installations carefully. Strive for simple, reliable, predictable. Gravity has yet to fail!