Sunday, May 29, 2011

Second Leak Fixed

The day started with spraying the white primer for the paint coat on the last stack of parts before the fuel tank. That was quickly done and the parts were left for drying.



Warning from the future!
Here I am traveling back in time from 2.5 years after I created this easter-egg to warn you NOT to follow my steps in how I tested the fuel system for leaks. While I was successful in finding and eliminating the leaks this way, I also created a serious problem in the fuel system that could have caused quite some more pain than it did in my case (which was luck alone!). 
The water in the system will cause the fuel pump to rust and decrease its performance significantly (or make it fail). This will lead to low fuel pressure issues on long WOT runs and will require a fuel pump replacement to fix (I am not sure if also the transducer might have been damaged by the water).
My suggestion is to use air alone at this point and to ensure that the temperature is the same when comparing PSI values from over-night runs. The alternative is to use actual gas but do recommend against it as it is too dangerous to mess with gas when there is a safer option.

Then it was back to the pressure test. The thread sealant should have had enough time over night to cure enough to withstand a few pounds.... I set everything back up again and set the pressure to around 7 psi. It didn't show an immediate drop so I let it alone and installed the servos for the autopilot.


That was kind of awkward to install first I thought I had to break my hand to get the servo in position with the pushrod already installed, then getting at the hex nuts to screw the servo in place was hard too. My hand seems a bit on the large side compared to the average Oregonian. I finally got all in only to realize I had forgotten to put the Loctite on. Argh! All out and back in. The second time was easier it appeared. Either my hand shrunk or I got used to the tight space. Anyway, I think installing the pushrod on the servo arm AFTER the servo has been installed would have been easier.

Back to checking the pressure.

Ouch!

It dropped half a pound while I was installing servo one. What the hell is this?

I pulled my soap water gun out and threatened those fittings but it didn't help. The leak must be a small one and hard to see with soap water. I used the flooding method and finally found another fitting that built up a foam collar although much slower and less impressive as the one I found yesterday. This time it was the forward fitting on the fuel valve. I drilled out the rivets holding the valve assembly to better work on it, removed the fitting and cleaned the threads as well as I could. Washed everything off with alcohol and smothered it with Rectorseal No.7 and put it back in. I gave it about 15 minutes to cure as I felt like I really had it with this cr*p. I pressure tested the valve assembly while still on the work bench and confirmed it was good to operating pressure of the fuel pump (about 5 psi). The difference was that I must have been worried about stripping the threads last time I put it in. Only the first 5 threads where in the valve and now I've screwed it almost completely in.
I put it back in the line but didn't want to give it my usual 8 psi test yet as the sealant was all fresh and I didn't want to blow it out with too early of a test. So I worked on the second servo.


(Comment from the past: You see this roll servo neatly installed at this point. DO NOT DO THIS! This is too early as it will prevent you from inserting the very long flaperon pushrods through that opening in the aft bulkhead that is blocked by the servo. Wait with this step until the pushrods are in place.)

Much easier to install although I don't know how to torque the forward two nuts. There's no way you can put a socket on them AND get a torque wrench connected to it. I torqued by feel instead.

Back to the pressure test. I filled the system with 8 psi and waited a little. I noticed a very small but noticeable drop.

Jeeeez!

What else can I do? Oh, I know! No more air! I need a liquid to see were it drips. I decided to hook up a small pump I have to fill the system with water (while the pump is on you have to push the gascolator drain to allow the air to escape).



Once the system is filled with water, I squeezed the water hose to stop the back bleeding and the I could add water through the Schrader valve to set it to the pressure desired. My gauge, Schrader valve assembly is above the water line, so this is the only part of the system that is not flooded.


Back to 8 psi again and waiting.

I went in the paint booth to check the parts and they were dry enough to get sanded with SB pads. So that's how I spent the waiting time, getting the parts ready for painting.

After half an hour I checked the gauge again.


I grabbed the work light and checked for leaks in the tunnel. Everything was perfectly dry. Not even moisture.
I sprayed the parts above the water line with soap again and couldn't see a blip. I checked the drain valve, not even moisture.
At this point I was considering selling the project.

Breath, Breath, Think!

I checked the water hose.


And there it was! A few drops of water were coming out of the large hose and resting on the clear line. As you can see I had a diameter issue here and no converters. This hose-in-a-hose thing with a clamp was all I could come up with. And that's were it leaked.
I clamped it a bit more and in different places and finally saw the bleeding stop and the pressure staying where it was.

Now the only unanswered question is: Why does it work with water but not with air? It looked like I had a very slow drop before doing the water test and I didn't have that hose hooked up then. I had a plug sitting on that end instead.

So was this test successful? Hm, I don't want fuel leaking into my plane. If I'd be flying with water I think I'd be good now. Unfortunately, that's not the case, is it?

No comments:

Post a Comment