What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Fuel Tank Leak Test Question

PhatRV

Well Known Member
I am in the process performing the leak test of my RV8 fuel tanks using the Vans balloon method. So how easy it it to detect the leaks using the soapy water? I shined the bright light on the rivet lines but didn't see anything bubbling other than the bubbles created by the spray nozzle. How long should I wait for to check for the balloon deflate, it if deflates as time goes by?

Thanks
 
I am in the process performing the leak test of my RV8 fuel tanks using the Vans balloon method. So how easy it it to detect the leaks using the soapy water? I shined the bright light on the rivet lines but didn't see anything bubbling other than the bubbles created by the spray nozzle. How long should I wait for to check for the balloon deflate, it if deflates as time goes by?

Thanks

Watching the balloon inflate or deflate over time is useless unless the temperature and ambient air pressure are completely stable. The balloon is to apply modest pressure to the tank so you can do the soap bubble test.

Take a good look at the joints between the skins and ribs and the skins and aft baffle. Take your time on this so you don't miss anything. It is a pain to fix a tank leak at this point, but once the tank is painted, it'll go from painful to heartbreaking.
 
Immediate

I am in the process performing the leak test of my RV8 fuel tanks using the Vans balloon method. So how easy it it to detect the leaks using the soapy water? I shined the bright light on the rivet lines but didn't see anything bubbling other than the bubbles created by the spray nozzle. How long should I wait for to check for the balloon deflate, it if deflates as time goes by?

Thanks

If there is a leak with air, the bubbles will grow immediately. The bubble test is super sensitive. As long as there are bubbles over every inch of the seam at some time, and the bubbles are not growing, you should be good. It is important that you put bubbles on the seam in question, and not just liquid water.
 
If you want to see what a "positive" soap bubble leak test looks like, try this: Get an extra balloon and inflate it. Place a piece of scotch tape on the balloon. Poke a pin through the piece of tape and the balloon (the tape prevents the balloon from popping). Now spray your soapy water on your newly leakified balloon. You'll see immediately what a leak looks like - it'll be a bona fide bubble factory!
 
I'm not a big fan of the balloon test. Most balloons gradually lose air even with no leaks. I make a homemade manometer (just a loop of clear tubing partially filled with water). The pressure will change up and down as the room temperature changes but I expect the tank to hold pressure several days.

As a final test I fill the tank 1/2 - 2/3 full with 100LL. Turn the tank as necessary to expose everything to the fuel. There shouldn't be any leaks after the manometer test but the dye will make it obvious if there are.
 
SNIP....

As a final test I fill the tank 1/2 - 2/3 full with 100LL. Turn the tank as necessary to expose everything to the fuel. There shouldn't be any leaks after the manometer test but the dye will make it obvious if there are.

+1 on this. The ballon leak test is good for obvious leaks, but not for finding those irritating ?weeping rivets? that are good for showing up just after you paint.

After I?m done building the wings I take the tanks off, fill them perhaps halfway with 100LL then let them sit for several weeks in different positions. I put paper towels under whatever side is down. When ready to mount back on the wings I slosh the gas in the tanks and drain via the sump through a plastic gas filter, and repeat as needed. This last step gets whatever junk out of the tanks before it ends up in your fuel filter.

Carl
 
As a final test I fill the tank 1/2 - 2/3 full with 100LL. Turn the tank as necessary to expose everything to the fuel. There shouldn't be any leaks after the manometer test but the dye will make it obvious if there are.

The leak test on the two tanks using the balloon method showed no obvious leak since no extra bubble was found. I didn't trust the result yet. I slowly unscrewed the schrader valve at the drain plug and I could see a lot of new bubbles forming. So at least the balloon test passed.

However, I will test it with fuel as you suggested. I don't have access to 100LL but I can test it using regular unleaded. Hopefully I can place the tank on top the paper towels and they can point to weeping leaks without the aid of the blue dye.

Thanks.
 
I think it'll be hard to use regular auto fuel without any coloring. If you can't get 100LL you might try one of the UV fuel dye kits or something like that.
 
Auto fuel (at least the E-free stuff) isn't colorless; it stains brown instead of blue. Trust me; I know.....

I wholeheartedly agree on using a manometer. If there are no leaks, the water level will move up/down with temps, but will never continue down unless there's a leak. Having said that, it can be really difficult to get the fill cap to seal completely. I've resorted to duct tape when I'm in a hurry & don't have o-rings on hand to get the cap sealing perfectly.
 
Auto fuel (at least the E-free stuff) isn't colorless; it stains brown instead of blue. Trust me; I know.....

I wholeheartedly agree on using a manometer. If there are no leaks, the water level will move up/down with temps, but will never continue down unless there's a leak. Having said that, it can be really difficult to get the fill cap to seal completely. I've resorted to duct tape when I'm in a hurry & don't have o-rings on hand to get the cap sealing perfectly.

True, auto fuel will make a stain, but it takes a lot more of it for a much longer period of time, for it to show.

The difficulty in sealing every single leak path (fuel cap, etc.) is why the balloon / bubble test is recommended.
If bubbles are seen at locations that aren't a concern, such as the fuel cap or where the balloon is attached, they can be ignored.

To clarify confusion that seems common.... the recommended use of the balloon has never been to monitor the balloon to determine if a leak is present.
It's only purpose is to be a simple way to pressurize the tank to a very low pressure value without having a gauge that is accurate to read only 2 PSI.
 
Last edited:
Thanks to everyone who replied. After confirming that there was no bubbling on both of the tanks, I called Vans support and they commented to me that if there is no bubbling, the tank should be leak free, and the deflation of the balloon isn't a good indicator, as in the fuel tank cap leaking.

BTW, before I Proseal the baffle, I tested both tanks using plain water. The fuel caps always had tiny amount of water weeping through the o-ring. When I tested the tank last night, I coated the fuel receptacle with plenty of Vaseline and that created an air tight fitting when the fuel caps were locked. With the combination of Vaseline and the Gorilla tape, there was no soapy bubbling.

An interesting thing that happened this afternoon. Since I decided to leave the balloon on the right wing tank just to see how long the balloon would hold, the warming noon temperature in the garage heated the air inside the tank and inflated the balloon even more. While working on other parts of the wing, I heard a loud pop. It was the balloon that pop! This event alone increased my confidence the tanks.
 
An interesting thing that happened this afternoon. Since I decided to leave the balloon on the right wing tank just to see how long the balloon would hold, the warming noon temperature in the garage heated the air inside the tank and inflated the balloon even more. While working on other parts of the wing, I heard a loud pop. It was the balloon that pop! This event alone increased my confidence the tanks.

Reading this makes me think you had your tank pressurized way beyond what is intended for the pressure test.

Assuming that others may be doing the same I will describe what is intended (which I thought was well detailed in the instructions but maybe not).

To avoid potentially damaging the tank (or maybe causing leaks that wouldn't have been there otherwise), the test pressure needs to be no more than about 2-3 PSI.

To attain this pressure, the balloon should only be inflated just enough for it to develop inflated shape (not be limp) and then grow in size very slightly (maybe 1/2" bigger in cross section dimension).
It should not come anywhere close to being inflated to the size that you would if it was being used as a party balloon.
 
Pressure

Reading this makes me think you had your tank pressurized way beyond what is intended for the pressure test.

Assuming that others may be doing the same I will describe what is intended (which I thought was well detailed in the instructions but maybe not).

To avoid potentially damaging the tank (or maybe causing leaks that wouldn't have been there otherwise), the test pressure needs to be no more than about 2-3 PSI.

To attain this pressure, the balloon should only be inflated just enough for it to develop inflated shape (not be limp) and then grow in size very slightly (maybe 1/2" bigger in cross section dimension).
It should not come anywhere close to being inflated to the size that you would if it was being used as a party balloon.

2-3 Psi? VAF is not Vans but every thread I reviewed on VAF before testing my tanks suggested 1 psi was more than enough. I used a home brew manometer to 27" and left them a month. Other than temp change, the manometer never moved. Now I'm wondering if they will hold at 2-3 psi and why the high pressure.
The 7 manual refers testing to the attached instructions using the Vans test kit.
YEyOOEel.jpg
 
Last edited:
Oops, it?s too late now. I got the right tank on the wing with all the bolts and screws torqued. Can?t do much now other than mounting the left wing. The only consolation is that it was the same balloon used on the left tank. It was in the stretched condition when I tested the right tank so the pressure should not be as high as if it is inflated right out of the bag.
 
2-3 Psi? VAF is not Vans but every thread I reviewed on VAF before testing my tanks suggested 1 psi was more than enough. I used a home brew manometer to 27" and left them a month. Other than temp change, the manometer never moved. Now I'm wondering if they will hold at 2-3 psi and why the high pressure.
The 7 manual refers testing to the attached instructions using the Vans test kit.
YEyOOEel.jpg

1 PSI is enough.

Note that I said no more than 2-3 PSI. I didn't say Use 2-3 PSI. The point I was trying to make but apparently failed to was that "Anything more than 2-3 PSI could damage the tank".
 
Pressure

1 PSI is enough.

Note that I said no more than 2-3 PSI. I didn't say Use 2-3 PSI. The point I was trying to make but apparently failed to was that "Anything more than 2-3 PSI could damage the tank".

Sorry Scott. I wasn't calling you out. I just remember my tank groaning and making horrible noises up to 1psi and wondered if a tank could handle 2-3.
 
Left Tank Retested: No Leak

All this talk about over pressure using the balloon could damage the tank got me worried. Since the left tank is still unmounted, I went ahead and retested it. This left tank saw almost the same amount of pressure as the right tank because the balloon was inflated at about the same size.

There was NO leak. Yeah. :)

I think the right tank, which has been installed on the wing, is fine too. I will retest it installed on the wing which mean I won't be able to look at baffle and the inner rib. It will be a warm fuzzy test only.
 
My tank tested OK 6 years ago when I built it. I used Mogas a couple of years ago and developed leaking rivets above the fuel line. Fixed them with Loctite green, but now I am suddenly getting more leaks. Fuel pouring out of the top corner of the tank on the inboard end. No leakage unless tank has over 35 litres.
After a couple of attempts to fix the problem, which seems to be getting worse, I have removed tank and cover plate. The proseal internally looks OK but it seems soft enough to peel away with a finger nail, the proseal externally at the upper corner is soft and that is where the leak is coming from. has anyone experienced proseal softening with age or exposure to mogas.
I have done a bodgyjob of sealing the corner up and will next apply pressure and do a soap bubble test. It rather looks as if I am up for either rebuilding a tank or building a new one. Not what I really want to do. If it is only the leak in the corner I suppose I could apply more proseal internally and hope that works.
Does anyone have any other ideas?
 
There is another thread talking about Proseal turning into mush when the tank is leaking. This gets me worried because I thought the Proseal is fuel proof and the leaking is caused by improper sealing instead of Proseal disintegrating over time. Were you using the 10:1 mix ratio in your build or did you vary the mix ratio to compensate for the temperature?
 
There is another thread talking about Proseal turning into mush when the tank is leaking. This gets me worried because I thought the Proseal is fuel proof and the leaking is caused by improper sealing instead of Proseal disintegrating over time. Were you using the 10:1 mix ratio in your build or did you vary the mix ratio to compensate for the temperature?

We know it happens. We don't yet know why it happens, not in the chemical sense. It is not mix ratio; the manganese dioxide is a catalyst, not a component. Full reversion appears to be linked to evaporation of fuel in free air. Some older technical papers suggest a link to the formation of peroxides in fuel.

Most would be surprised to learn there is no mil spec or standard specific to polysulfide sealant use with avgas or mogas. All tests are done with JRF, a standard fluid emulating jet fuel. Does it matter? Clearly the sealant works under normal circumstances, inside the tanks. Avgas is standardized. Mogas is always a wild card.

Entirely aside, I'm not aware of any manufacturer's instruction which suggests mix ratio should be varied for temperature.
 
Thanks for the info. I don't vary the mix but other mentioned they used a hotter mix while working in colder climate.
 
Tank test

I used the bubble test and it was evident where leaks appeared to be. However, it did not get all of them. I highly recommend the fuel test. Put the tank on concrete. The concrete will stain or just be wet. Weeping rivets did not bubble test well. I learned the hard way after installing and then having to remove my tanks. Find some 100 LL somewhere. Not to hard.
 
Nobody mentioned Snoop. Cheap at about 4 dollars a bottle. Super sensitive to slightest leak. Has a little glycerine in it with the soap. Get it at a swaglok distributor most towns have one. If it doesn?t bubble with Snoop there is no leak.
 
I was thinking about the comment about using snoop. My initial thought, was cool, a better way to flush out bubbles. My concern would be what if there was a leak? Would the ingredients in snoop interfere with bonding new sealant to the existing sealant? Yes, you can clean the surfaces, but in this case I would be hesitant to mess with what has always worked.
 
Follow up on my problem. I am sure the leak was caused by mogas as the tank which only ever had avgas still does not leak. My bodgy job didn't work, so I removed the tank again and found a leak at the rear of the wing closest to the fuselage, where rib, top skin and aft wall of the tank meet. Much probing and I managed to find the leak, right in the corner. Eventually I got a 3/32 drill right through and with much probing and scraping was able to fix the problem with proseal. The cost of proseal was less than the cost of freight. So far all seems well, It tested OK and has been filled twice since repairing it.
 
Back
Top