stevemcgirr
Well Known Member
Allow me a short preamble before I make a few observations on checking for leaks in your tanks. Both of my tanks had passed the Van?s ?balloon test? with seemingly no leakage of air, long ago when the wings were being built in my garage. There were no apparent leaks when I filled five gallons per side to do the fuel flow tests, nor even during the slow speed taxi tests. So it was frustrating about a month and a half ago, when I first completely filled the tanks to do the capacitance sender/EFIS calibrations, to find that my right tank had developed a leak.
I first noted a trail of blue coming from the outboard side. I determined that the tank would have to be pulled off the wing to properly repair, and found the leak, due to the trail of blue, coinciding with an area of soap bubbling on a seam on the outboard side. I repaired the tank by cutting an access panel and applying tank sealant in the usual way. After sealing the access panel on, and allowing a week for cure, I did another air challenge test to 14 inches of water on a manometer I built (which I had used to check calibration of pitot and static lines). Overnight (24 hours later) there was a column 12 inches in height. Since I could find no soap bubbles, I figured that I was probably seeing ?acceptable? leakage from the hoses I was using for testing and re-installed the tank. I re-filled to full. No leak, seemingly, overnight in the cold hangar. Next day though, when I brought the plane out in to the gradually warming sun, a new leak became apparent and intensified, though the origin was hard to determine. Off with the tank AGAIN. (in 53 minutes, though this wasn?t a skill I had hoped to perfect). This time there was no obvious trail of blue. When I pressurized again to a 12 inch column of air, I could see no movement in the manometer. I could find any soap bubbling, either. So, since my wife was still at work, I took the tank home for the ?hot tub test?. Submerging about a quarter of the tank at a time and pressurizing with as much wind as I could blow, I found an intermittent leak at the corner seam at the wing root, above the access panel for the fuel pickup. Moving the overhang of aluminum here caused the leak to be more obvious. The second area of leakage has now been sealed, and is curing.
Now to the observations and the musings. I had been of the understanding that to do an adequate ?leak test? that you sealed off all but one outlet on the tank, blew the tank up with air sufficient to blow up a balloon, and waited overnight to see if the balloon was still inflated. If it was, you could feel confident that you would have no leaks. I was curious though as to how much of a challenge this actually was. What pressure does a balloon exert, after all? What pressure SHOULD we be checking for? How much is TOO much?
It turns out that the balloons I have access to, which inflate to approximately a foot in diameter, require about 20 inches of water pressure to start inflating or about .72 psi (I measured). Once inflated, and this varies depending on the age of the balloon and how many times it has been inflated, the balloon exerts about 5 inches of water, or about .18 psi at steady state. This would not have been enough to sense my leak. The average male can produce a peak of 39-40 inches of water pressure, or about 1.45 psi (citation: www.ncbi.nlm.nih.gov/pmc/articles/PMC1501025 ). So while you can inflate most balloons, you can?t hurt your fuel tank. I read somewhere that it was recommended to inflate tanks during the leak check to ?about 3 psi?. If this is the case, you can?t do it with a balloon. To put this in another perspective, if your vent line were blocked, and you went from sea level to 20,000 feet, ignoring temperature changes for a minute to make the math easier, the pressure change would be 20 inches of mercury (standard lapse rate: one inch per thousand feet) corresponding to a tank pressure change of 9.8 psi, so I guess that would be a worst case scenario for tank pressurization.
The take home messages for me are that I will check my tank with a manometer, rather than with a balloon. I will pressurize to about 3 psi (which is a column of water 8 feet high!). I will probably leave things overnight (though the manometer is very sensitive, and you can see movement in the meniscus quickly if there is a leak). And I have made very sure that the fuel vent lines are open and unobstructed. With a functioning vent line, the most pressure a tank should see is if a column of fuel gets pushed up into the vertical section of the vent up to the longeron, a height of less than 18 inches, so less than .7 psi. Overkill perhaps, but I REALLY don?t want to have to take the tank off a third time.
I first noted a trail of blue coming from the outboard side. I determined that the tank would have to be pulled off the wing to properly repair, and found the leak, due to the trail of blue, coinciding with an area of soap bubbling on a seam on the outboard side. I repaired the tank by cutting an access panel and applying tank sealant in the usual way. After sealing the access panel on, and allowing a week for cure, I did another air challenge test to 14 inches of water on a manometer I built (which I had used to check calibration of pitot and static lines). Overnight (24 hours later) there was a column 12 inches in height. Since I could find no soap bubbles, I figured that I was probably seeing ?acceptable? leakage from the hoses I was using for testing and re-installed the tank. I re-filled to full. No leak, seemingly, overnight in the cold hangar. Next day though, when I brought the plane out in to the gradually warming sun, a new leak became apparent and intensified, though the origin was hard to determine. Off with the tank AGAIN. (in 53 minutes, though this wasn?t a skill I had hoped to perfect). This time there was no obvious trail of blue. When I pressurized again to a 12 inch column of air, I could see no movement in the manometer. I could find any soap bubbling, either. So, since my wife was still at work, I took the tank home for the ?hot tub test?. Submerging about a quarter of the tank at a time and pressurizing with as much wind as I could blow, I found an intermittent leak at the corner seam at the wing root, above the access panel for the fuel pickup. Moving the overhang of aluminum here caused the leak to be more obvious. The second area of leakage has now been sealed, and is curing.
Now to the observations and the musings. I had been of the understanding that to do an adequate ?leak test? that you sealed off all but one outlet on the tank, blew the tank up with air sufficient to blow up a balloon, and waited overnight to see if the balloon was still inflated. If it was, you could feel confident that you would have no leaks. I was curious though as to how much of a challenge this actually was. What pressure does a balloon exert, after all? What pressure SHOULD we be checking for? How much is TOO much?
It turns out that the balloons I have access to, which inflate to approximately a foot in diameter, require about 20 inches of water pressure to start inflating or about .72 psi (I measured). Once inflated, and this varies depending on the age of the balloon and how many times it has been inflated, the balloon exerts about 5 inches of water, or about .18 psi at steady state. This would not have been enough to sense my leak. The average male can produce a peak of 39-40 inches of water pressure, or about 1.45 psi (citation: www.ncbi.nlm.nih.gov/pmc/articles/PMC1501025 ). So while you can inflate most balloons, you can?t hurt your fuel tank. I read somewhere that it was recommended to inflate tanks during the leak check to ?about 3 psi?. If this is the case, you can?t do it with a balloon. To put this in another perspective, if your vent line were blocked, and you went from sea level to 20,000 feet, ignoring temperature changes for a minute to make the math easier, the pressure change would be 20 inches of mercury (standard lapse rate: one inch per thousand feet) corresponding to a tank pressure change of 9.8 psi, so I guess that would be a worst case scenario for tank pressurization.
The take home messages for me are that I will check my tank with a manometer, rather than with a balloon. I will pressurize to about 3 psi (which is a column of water 8 feet high!). I will probably leave things overnight (though the manometer is very sensitive, and you can see movement in the meniscus quickly if there is a leak). And I have made very sure that the fuel vent lines are open and unobstructed. With a functioning vent line, the most pressure a tank should see is if a column of fuel gets pushed up into the vertical section of the vent up to the longeron, a height of less than 18 inches, so less than .7 psi. Overkill perhaps, but I REALLY don?t want to have to take the tank off a third time.
. (I guess a reverse leak, but only to a point!)
