18-09: Fuel tank Testing

 So its been a week or so and I decided to follow the plans to test the tanks. It is a pretty simple process: on the vent line you place a rubber tube held in place with a worm drive clamp and the other end of this tube has a schraeder valve (bike valve) installed. On the fuel pickup fitting you cable tie a balloon and it is simply a matter of pumping air into the tanks using a bike pump until the balloon inflates. A successful test is proven by the balloon staying inflated and there being no air leaks (hissing sounds) and no bubbles anywhere when you spray on soapy water. 

A couple of comments on this procedure straight up - you need to seal the rubber tube with thread tape on all connections - i know this is not permitted in a fuel tank setup but i was very careful not to get any tape near the ends of the fittings. Without the thread tape, everything just leaks. 

This is the rubber tube / schraeder valve installed onto the fuel vent fitting.

Thread tape is used to get a good seal

and a balloon is placed on the fuel pickup fitting.

I used some packing tape over the tank caps just in case, but i also tested the tanks with this removed and the caps did not leak any air.

Right Tank - Balloon Test

I started with the setup above for the right tank and found no leaks and the balloon stayed inflated for over 12 hours - SUCCESS!! - no leaks in the right tank. You can see in the video below the balloon changing size due to local temperature and pressure changes.

The soap made a real mess, so the tank was removed to the lawn and washed down.

Here you can see both tanks maintained a full balloon so were seemingly ok!

Left Tank - Manometer Test

During the setup for the left tank i was able to get the balloon to inflate for the test (as seen in the image above), and there were no bubbles etc but i could not get the balloon to stay blown up for the night - deflated (see what i did there) i started checking for leaks and found that the balloon connection was leaking. So i decided to use a Water Manometer to test the tanks instead of a balloon - this seemed a lot more scientific. 

The setup for the manometer was similar to the balloon, however instead of a balloon a clear tube was placed onto the fuel pickup and closed tight using a worm drive clamp (instead of a cable tie). The tube is snaked down and then up again such that there is room for an amount of water in the tube to be pushed down 14" on one side, then lifted 14" on the other. The combined 28" of water being lifted equates to a little over 1psi - the requisite pressure called out by vans. 

Here you can see the setup - as i pressurise the tank, the pressure inside is acting upon the water, which is being acted upon on the opposite end by both gravity and atmospheric pressure. By pushing the water down 14" on one side and up 14" on the other, this is equivalent to lifting 28" of water - and equates to a pressure a little over 1psi differential between atmospheric and inside the tank. 

The manometer is very sensitive - mostly to temperature changes. As you can see in the below video the left tank was setup with the manometer attached and the water level changes quite a lot over time. 

I came out to the shed in the morning horrified to see the water level had dropped! But so had the temperature. So i did not touch the setup and i waited until the temperature in the shop rose to a similar level - and the water level rose up past the line as the temperature increased above the original temp. 

This result, coupled with the lack of bubbles meant that i had a good left tank - NO LEAKS!

Note- in the videos below the “airspeed indicator” is actually a thermometer! 

Right Tank - testing again using the manometer

I liked the results of the manometer a lot more than the balloon test method - it seemed so much more scientific. So i setup the right tank with the same system, pumped it up and left it over night. I was completely expecting it to be fine, as the balloon on this tank stayed inflated. 

The next morning i noticed that the water level had also dropped - by almost twice as much. Looking at the temperature changes compared to the left tank, i thought this might have been a little too much of a drop. So to test my theory, i pumped the tank back up to 14" and decided to observe what happened. To my horror the pressure was dropping quite a bit in the tank over a very short period of time. Note how the water level is dropping but the temp is staying constant - there was no real change in the pressure at my local airport - so I’ve got a problem. 

In order to find where this leak was, i again pumped up the tank as far as the tube would allow (perhaps 1.3 PSI) and went searching with the soapy water. To my horror i found the following:

Bubbles are bad

I recall when i was installing the right baffle and was dropping it down on onto the proseal beads, that i noticed it was hanging maybe 1/4" too far inboard, so i pushed it outboard and clecod it in place. Perhaps this pulled the sealant away from the end where the rib flange and the baffle web meet? 

I pulled out the old crappy borescope and had a look. As you can see in this video there is a small circular section where there is no proseal up against the rib flange intersection - somehow this must be where the air is leaking out (as it is in approx the correct location to the bubbles on the outside). 

I think (hope) this is the offending location. It matches to where the bubbles were on the outside. 

I covered the end rib flange in a thin smear of sealant, but perhaps these ribs moved apart slightly when riveted together creating a leak path. Who knows - but it needs to be fixed. 

Looks like i have joined 'got a leak that needs fixing' group - at least i found it at this early stage.