Re: REFLECTOR: Fuel Drains

["William Wade" <chipmunk@uninets.net>]

   Al- I hadn't considered the crossflow between the tanks- one of several good points you made. Hopefully that would be enough to get the water into the header. Otherwise I would think that replacing any significant amount of header fuel by water would take a long time. If both tanks were level the displaced fuel would have to go to the tank with the water, creating a backflow. The 1/4" outboard line might provide a return path though. When using water levels I have had bubbles trapped inside 1/2" hose for hours. 3/8" tubing has a much smaller area and there would be less differential in specific gravity between water and gas. However there's no guarantee the tank with the water would draw down first when the engine was operated. That seems to vary with circumstance. My empirical thoughts. Anyone want to run a system test with colored water?

  My system can handle about 12 oz. of water before the header pickup is reached. If 3 gallons was introduced the header would eventually be filled completely. Hopefully that would be on the ground. Point is I thought about strake drains and couldn't see how to make them work within the shape of the strake. Maybe I should have tried harder.  -Bill

 

prototype 'Super' Chipmunk N18EF
Velocity Classic RG N6098S in the works

 

First, the assumption is being made that water will not be able to find its way down a 3/8" line that has fuel in it without flow. Has this been shown to be true?

 

 Any amount of flow would cause the water to go through the line to the sump tank.  When you add fuel to one tank; the level differential between tanks will cause flow to/through the sump tank; thereby trapping the water in the bottom of the sump.  Always fill the second tank to a level higher than the first.  In topping off, this will happen whether you like it or not unless you're filling both tanks simultaneously.  And one might expect that if water is coming from the fuel source; it would come at the start of filling.