As an alternative to the balloon test for leak-testing fuel tanks, I strongly recommend the manometer test, and I am providing a link to an Excel spreadsheet that may help you be less nervous about the results.
Why not a balloon test? Well, standard rubber party balloons leak, and it can be difficult to attach them to a fitting so the fitting-balloon joint doesn't leak. Also, with a balloon, you really don't have a good idea of how much pressure you're putting into your tank. Suppose you put 1 psi into the tank, and your tank has a surface area of 1000 square inches. What's the total net pressure on the skin inside the tank? Answer: 1000 lbs. Now put 25 gallons of avgas in the tank. That weighs about 150 lbs. S'pose that 150 lbs is evenly distributed over half the area of the tank, e.g., the bottom half of the tank. That amounts to about 30% of the pressure per square inch that you would put on the skin-rib, skin-baffle, and baffle-rib joints doing your balloon test at 1 psi. So if you do a 1 psi balloon test, you're prying apart the joints with more than three times the pressure you'd have with a full load of fuel, assuming you stick to 1 g flight most of the time. Hope you scuffed the faying surfaces inside the tank!
So, how do you build a manometer, and how do you configure the tank for the manometer test?
The manometer I built is simply a loop of transparent tubing mounted on a board. Fill the tubing with water to about 1/2 the height of the loop. You can mix the water with food coloring to make it more photogenic/visible. I used an 9" x 27 x 0.75" board, which I think I salvaged from my wing spar crate. I made my manometer using 1/4 in. I.D. x 3/8 in. O.D. Clear Vinyl Tubing, readily available from your local big-box hardware store with the orange and white logo, and probably from their main competitor as well. I used cable staples to attach the tubing loop to the board. One side of the manometer tube connects to the tank as explained below, and the other side is open to the atmosphere.
My fuel tanks have an AN432-4D vent line fitting. I used a decent flaring tool (Rigid Model 377) to make a 37-degree flare on a 4" long piece of 1/4" OD 3003-O Versatube, which fits nicely inside the 1/4" ID clear vinyl tubing, to connect the fuel tank to the manometer. I used two small hose clamps to clamp the clear vinyl tubing to the aluminum tubing. I put a thin layer of EZ Turn (fuel lube) on the outside of the aluminum tubing to help achieve an airtight seal between the aluminum tubing and the vinyl tubing. I used the fuel lube very sparingly on the flare joints, too, not having any Del fitting seals at present.
My fuel tanks have a VA-261 fuel strainer for the fuel feed. I capped it with an AN929-6.
My fuel tanks have a CAV-110 fuel drain, AS p/n 05-01865. The Schrader valve Van's sells in their fuel tank test kit fits nicely in my tank's fuel drain fittings. Put some EZ Turn on the threads of the Schrader valve that screw into the tank drain port.
Unless you're within a 30-minute drive of an AS location, and you're building a -12 or -14, do yourself a favor and buy the Van's fuel tank test kit, https://store.vansaircraft.com/fuel-tank-leak-tester-fuel-tank-test-kit.html. It has both the AN929-6 cap and the Schrader valve.
So, now let's assume you have the Schrader valve installed in the tank drain port, the fuel strainer capped, the vent line connected to the manometer*, and the manometer filled about halfway with water. Now you need to connect a bike pump to the Schrader valve, and put some pressure in the tank. My 25 gallon tanks required about six strokes of my Harbor Freight $11 (on sale) bike pump to achieve a 10" difference between the low-side and high-side water levels in the manometer. I don't recommend going above a 15" water level differnce, which equatses to about 1/2 psi. Remember, your tank has a lot of square inches of surface area.
Once you've got the desired amount of pressure in the tank, disconnect the bike pump and put the cap on the Schrader valve.
A couple of things to be aware of:
(1) When you first pump air into the tank, the water level on the high side of the manometer will rise, then fall. Some of the fall may be due to cooling of the air as it moves from the pump to the tank. Some may be leak-back into the bike pump. Be patient, and adjust the pressure as needed.
(2) In my case, the biggest leak source was the fuel filler caps. The pressure inside the tank pushes the caps out, compromising the seal. I used a small amount of silicone faucet grease on the filler cap O-rings and tightened the nuts on the bottom of the caps to mitigate this problem. If you tighten the nut on the fuel cap, remember to loosen it to its original position when you're done testing, or it will be difficult to remove the cap and you'll deform the O-ring.
If the water level in the manometer is still changing at a rate you can detect by staring at the tube for a minute or so five minutes after you disconnected the pump, you either have a loose connection or a leak in your tank. A bottle of Oatey All-Purpose Leak Detector comes in very handy for figuring out where the air is escaping from the tank.
The water level in the manometer will fluctuate as the ambient atmospheric pressure changes and the temperature of the air inside the tank changes. (Remember PV=nRT?) At my elevation, 5400' MSL, a 1 deg. F change in the temperature inside the tank results in a water level change of about 0.65". (This is measured as the difference between the water level on the low side and the water level on the high side of the manometer.) An altimeter setting change of 0.1" of mercury results in a water level change of about 1.36".
Let the water level in the manometer stabilize for at least five minutes before taking the initial measurement of water level on the high (atmosphere) side minus water level on the low (tank) side.
Use this Excel spreadsheet to figure out what happens if the temperature and/or atmospheric pressure in your test environment changes:
Enter the station pressure at the beginning of the test in cell B4 of the spreadsheet. Note that this is the actual pressure in inches of mercury, not the altimeter setting you'd put in the Kollsman window of your altimeter. If you know the current altimeter setting and the elevation at your test site, you can use https://www.weather.gov/epz/wxcalc_stationpressure to calculate the station pressure at your test site.
Enter the temperature at the beginning of the test in cell B8.
Enter the relative humidity at the beginning of the test in cell B11.
When you want to see how much the water level should have changed due to a change in ambient pressure and/or a change in the temperature of the air inside the tank, enter the new station pressure in cell E4 and the new temperature in cell E20 of the spreadsheet. The water level change you should see if there are no leaks in your tank and fittings will be found in cell E29 of the spreadsheet.
Let everything sit undisturbed for at least 24 hours (preferably 48 hours) before taking a final measurement of high-side water level minus low-side water level. If the water level difference matches the spreadsheet prediction well, you can breathe a sigh of relief.
But if the water level difference drops much more than the spreadsheet prediction, grab your bottle of Oatey Leak Detector and start swabbing fittings and rivets and looking for bubbles. You may want to pump the tank up until you get about 15 inches of water level difference in the manometer tubing to accelerate the leak a bit.
If you find a leaking rivet, even in a place you can't reach through the filler cap flange or the fuel sender hole, it may be possible to fix the leak without cutting the tank open. Another builder in my local area had one leaky rivet near the aft end of a rib, where the shop head was hidden under the baffle flange of the rib. He mixed up a small batch of tank sealant and diluted it with toluene. He pulled a slight vacuum in his tank using a manual vacuum pump from a brake bleeder tool kit, using the manometer to monitor the level of vacuum he pulled. Then he dabbed some of the diluted tank sealant on the leaky rivet. He let the sealant cure for a few days and re-tested. No more leak.
Why not a balloon test? Well, standard rubber party balloons leak, and it can be difficult to attach them to a fitting so the fitting-balloon joint doesn't leak. Also, with a balloon, you really don't have a good idea of how much pressure you're putting into your tank. Suppose you put 1 psi into the tank, and your tank has a surface area of 1000 square inches. What's the total net pressure on the skin inside the tank? Answer: 1000 lbs. Now put 25 gallons of avgas in the tank. That weighs about 150 lbs. S'pose that 150 lbs is evenly distributed over half the area of the tank, e.g., the bottom half of the tank. That amounts to about 30% of the pressure per square inch that you would put on the skin-rib, skin-baffle, and baffle-rib joints doing your balloon test at 1 psi. So if you do a 1 psi balloon test, you're prying apart the joints with more than three times the pressure you'd have with a full load of fuel, assuming you stick to 1 g flight most of the time. Hope you scuffed the faying surfaces inside the tank!
So, how do you build a manometer, and how do you configure the tank for the manometer test?
The manometer I built is simply a loop of transparent tubing mounted on a board. Fill the tubing with water to about 1/2 the height of the loop. You can mix the water with food coloring to make it more photogenic/visible. I used an 9" x 27 x 0.75" board, which I think I salvaged from my wing spar crate. I made my manometer using 1/4 in. I.D. x 3/8 in. O.D. Clear Vinyl Tubing, readily available from your local big-box hardware store with the orange and white logo, and probably from their main competitor as well. I used cable staples to attach the tubing loop to the board. One side of the manometer tube connects to the tank as explained below, and the other side is open to the atmosphere.
My fuel tanks have an AN432-4D vent line fitting. I used a decent flaring tool (Rigid Model 377) to make a 37-degree flare on a 4" long piece of 1/4" OD 3003-O Versatube, which fits nicely inside the 1/4" ID clear vinyl tubing, to connect the fuel tank to the manometer. I used two small hose clamps to clamp the clear vinyl tubing to the aluminum tubing. I put a thin layer of EZ Turn (fuel lube) on the outside of the aluminum tubing to help achieve an airtight seal between the aluminum tubing and the vinyl tubing. I used the fuel lube very sparingly on the flare joints, too, not having any Del fitting seals at present.
My fuel tanks have a VA-261 fuel strainer for the fuel feed. I capped it with an AN929-6.
My fuel tanks have a CAV-110 fuel drain, AS p/n 05-01865. The Schrader valve Van's sells in their fuel tank test kit fits nicely in my tank's fuel drain fittings. Put some EZ Turn on the threads of the Schrader valve that screw into the tank drain port.
Unless you're within a 30-minute drive of an AS location, and you're building a -12 or -14, do yourself a favor and buy the Van's fuel tank test kit, https://store.vansaircraft.com/fuel-tank-leak-tester-fuel-tank-test-kit.html. It has both the AN929-6 cap and the Schrader valve.
So, now let's assume you have the Schrader valve installed in the tank drain port, the fuel strainer capped, the vent line connected to the manometer*, and the manometer filled about halfway with water. Now you need to connect a bike pump to the Schrader valve, and put some pressure in the tank. My 25 gallon tanks required about six strokes of my Harbor Freight $11 (on sale) bike pump to achieve a 10" difference between the low-side and high-side water levels in the manometer. I don't recommend going above a 15" water level differnce, which equatses to about 1/2 psi. Remember, your tank has a lot of square inches of surface area.
Once you've got the desired amount of pressure in the tank, disconnect the bike pump and put the cap on the Schrader valve.
A couple of things to be aware of:
(1) When you first pump air into the tank, the water level on the high side of the manometer will rise, then fall. Some of the fall may be due to cooling of the air as it moves from the pump to the tank. Some may be leak-back into the bike pump. Be patient, and adjust the pressure as needed.
(2) In my case, the biggest leak source was the fuel filler caps. The pressure inside the tank pushes the caps out, compromising the seal. I used a small amount of silicone faucet grease on the filler cap O-rings and tightened the nuts on the bottom of the caps to mitigate this problem. If you tighten the nut on the fuel cap, remember to loosen it to its original position when you're done testing, or it will be difficult to remove the cap and you'll deform the O-ring.
If the water level in the manometer is still changing at a rate you can detect by staring at the tube for a minute or so five minutes after you disconnected the pump, you either have a loose connection or a leak in your tank. A bottle of Oatey All-Purpose Leak Detector comes in very handy for figuring out where the air is escaping from the tank.
The water level in the manometer will fluctuate as the ambient atmospheric pressure changes and the temperature of the air inside the tank changes. (Remember PV=nRT?) At my elevation, 5400' MSL, a 1 deg. F change in the temperature inside the tank results in a water level change of about 0.65". (This is measured as the difference between the water level on the low side and the water level on the high side of the manometer.) An altimeter setting change of 0.1" of mercury results in a water level change of about 1.36".
Let the water level in the manometer stabilize for at least five minutes before taking the initial measurement of water level on the high (atmosphere) side minus water level on the low (tank) side.
Use this Excel spreadsheet to figure out what happens if the temperature and/or atmospheric pressure in your test environment changes:
Enter the station pressure at the beginning of the test in cell B4 of the spreadsheet. Note that this is the actual pressure in inches of mercury, not the altimeter setting you'd put in the Kollsman window of your altimeter. If you know the current altimeter setting and the elevation at your test site, you can use https://www.weather.gov/epz/wxcalc_stationpressure to calculate the station pressure at your test site.
Enter the temperature at the beginning of the test in cell B8.
Enter the relative humidity at the beginning of the test in cell B11.
When you want to see how much the water level should have changed due to a change in ambient pressure and/or a change in the temperature of the air inside the tank, enter the new station pressure in cell E4 and the new temperature in cell E20 of the spreadsheet. The water level change you should see if there are no leaks in your tank and fittings will be found in cell E29 of the spreadsheet.
Let everything sit undisturbed for at least 24 hours (preferably 48 hours) before taking a final measurement of high-side water level minus low-side water level. If the water level difference matches the spreadsheet prediction well, you can breathe a sigh of relief.
But if the water level difference drops much more than the spreadsheet prediction, grab your bottle of Oatey Leak Detector and start swabbing fittings and rivets and looking for bubbles. You may want to pump the tank up until you get about 15 inches of water level difference in the manometer tubing to accelerate the leak a bit.
If you find a leaking rivet, even in a place you can't reach through the filler cap flange or the fuel sender hole, it may be possible to fix the leak without cutting the tank open. Another builder in my local area had one leaky rivet near the aft end of a rib, where the shop head was hidden under the baffle flange of the rib. He mixed up a small batch of tank sealant and diluted it with toluene. He pulled a slight vacuum in his tank using a manual vacuum pump from a brake bleeder tool kit, using the manometer to monitor the level of vacuum he pulled. Then he dabbed some of the diluted tank sealant on the leaky rivet. He let the sealant cure for a few days and re-tested. No more leak.
