Fuel Tanks

Tonight I wanted to get the vent lines and fuel sender plates in place, and maybe the outboard end ribs if possible. I started out by inserting the snap bushings and routing the vent lines in both tanks. Next I installed the outboard sender plates (an operation which requires about five hands) and routed the wire from the outboard plate along the vent line run. The vent lines needed to be bent a touch at the outboard end to go through the clips under the gas caps, but that was no trouble. Once I had them roughly in place, I clecoed the inboard end rib in place and screwed the vent line to the bulkhead fitting. This finalized the location of the vent line, and I fixed the line in place by mixing up a bit of sealant and dabbing it around where the vent line goes through the gas cap clip.

Since I still had some sealant mixed up, I decided to go ahead and put the access plate on as well. I hadn’t scuffed these mating surfaces, so that was step 1, followed by a good acetone cleaning. For this application, I put a bead of sealant both on the rib and the access plate. I also swirled a bit of sealant on the tip of each screw before installing it, which gave me a nice sealant bead at the base of each screw. There shouldn’t be any issues with leaking around these guys:

A brief comic aside while I’m on the access covers: some of the nutplates had sealant built up in them, so I wanted to run a #40 drill through them to break up the buildup. When drilling like this, I’m used to holding the drill bit lightly with the fingers of my left hand to help guide it. Turns out that doing this wile wearing nitrile gloves can result in a pretty amusing tangle situation:

Finally, here’s a shot of one of the outboard sender plates:

I didn’t install the inboard plate yet because I’ll have to solder and crimp the ring terminal in place after I run the wire from the inboard end rib. That should be lots of fun. Tomorrow I’m going to shoot for getting the end ribs installed and wet riveted. I read an interesting idea for an early leak test, wherein you fill the tanks with water prior to installing the baffles. It’s not a perfect test, since fuel is less dense than water and can probably seep through bad spots better than water, but it’s still a way to maybe identify trouble spots and fix them before the tank’s closed – at which point fixing leaks might require creating some new access plates in the baffle. Thing is, the sealant needs to be decently cured first. I’m hoping that if I get everything riveted tomorrow night, then three days of sitting in a warm Houston garage will cure the sealant enough for me to do the water test on Saturday.

Well, the main thing I learned/concluded tonight is that I won’t be closing the tanks this weekend – at least not both of them. I think the best-case scenario is that the right tank might be ready for the water test Saturday or Sunday. This isn’t because I broke anything, just that I once again overestimated how much I can get done in the evenings after work.

I did realize earlier today that I made a mistake though – when I installed the outboard fuel sender plates last night, I forgot the little nylon tubing spacers that surround the screws to ensure they’re electrically isolated from the ribs. So first thing when I got out today, I got to put those in. At least I was able to do it without totally removing the plates; I was able to remove one screw at a time, insert the spacer, and then reinstall the screw. Bit tedious, but far better than completely removing and reinstalling the plates.

Next, it was time to address the end ribs. I was already beginning to suspect at this point that my goal of installing and riveting all the end ribs tonight was horribly optimistic, so I reduced my goal to only doing the end ribs for one tank. I was still intending, though, to just wet seal the rivets instead of doing the whole fay sealing thing – but that didn’t go so well. It didn’t take long to put a bead of sealant on each rib and get them in place, although getting them align was a bit of a pain. I put a cleco in every third hole or so, intending to rivet in between and then remove the clecos. I even got so far as to dab extra sealant into the dimples where needed, but when I tried to put the rivets in their holes, I had a fair amount of trouble. I don’t know if the holes are a little undersized or slight misaligned or what, but they’re going to need a little work before I can rivet.

It was getting late at this point anyway, so I removed all the rivets I’d tried to insert and reverted to the fay seal method. I 100% clecoed, and then went to work smoothing out a good fillet with the squeezed-out sealant. Cleanup was a little more fun for several reasons: first, the way these ribs go in kind of smears sealant around anyway, and I also wanted to clean up all the excess sealant from the outboard tank skin, outside of the rib. This area will overlap the joint plate by the leading edge, and I don’t want any sealant buildup here to affect the fit there.

So I guess tomorrow night I’ll clean up the holes as necessary and wet set the rivets. If I can get the rest of the sealing in the right tank done tomorrow, then I’ll have three days of curing in by Sunday, so maybe I can do the water test then. Even that might be a stretch though – in addition to setting the rivets, I’ll have to finish the wiring for the fuel senders, which has to be done inside the tank itself, and that terminal has to get sealed once it’s finalized. At this point, I should probably just resign myself to another weekend going by before I’m ready to close these things out. I know I’m just dealing with arbitrary goals I’m setting rather than true setbacks, but boy, does it ever feel like sealing these things is an endless job.

Here’s the inboard rib in place with the sealant and so forth:

The more time I spend on this stuff, the funnier it is to think I once imagined I’d do the end ribs of both tanks, plus the internal wiring and plumbing, all in one evening after work. Fortunately, though, the right tank is almost there – all that’s left to do is to install the inboard sender plate and finalize the wiring and so forth.
Nothing really complicated went down tonight, just sort of rote work. I pulled out around 2/3 of the clecos, cleaned out the holes, hit the dimples with a countersink bit, cleaned, dabbed in sealant, and squeezed the rivets, then repeated with the remaining rivets. It was a nice change to just use the squeezer on these instead of the rivet gun. Next up was the forward tank attach bracket; I just smeared sealant onto the contacting surface with a popsicle stick like I was icing a cake, then clecoed it in place. I recruited Josie’s help to bang those six rivets, since it was a bit of a stretch to both shoot and buck while working at the extreme front of the tank.
From there, I installed the flop tube, the remaining anti-hangup guide, and the little trapdoor, adding sealant strategically to help hold a few items in place. (I think I’ve come to the point of thinking that sealant makes everything better. Some kind of weird Stockholm Syndrome or something.)
I would have liked to have done the sender plate tonight as well, but I checked the time after doing the trapdoor and it was already midnight! Time to turn in for the night for sure. I think my new goal is to try to have both tanks done this weekend, with the exception of the baffles. That’ll give the sealant a week or more to cure, and I can do the gross water leak test next weekend before finally closing these things out.
Anyway, it’s photo time. Here’s a tank with no clecos:

Close-up of the forward attach bracket with the flop tube bulkhead fitting going through it:

Wide view of the inboard bay, where all the crazy stuff is at:

Close-up of the trapdoor and outboard anti-hangup guide:

And the flop tube bulkhead fitting from the inside:

Whew! Now to see if I can get the sender in place tomorrow night. Or maybe I’ll go ahead and do the end ribs for the left tank tomorrow, and save the sender plates on both tanks for a single sort of “final stuff” session.

Well, not a lot to report tonight. We had things going on in the morning, and then I may have taken a nap when we got home, but I finally did get out later in the evening to get the end ribs sealed and clecoed in place. The fun part about this task isn’t so much the sealing, but the cleanup afterwards. I’ve been more liberal with the sealant on these ribs, since in my mind they’re more critical then the interior ribs. As a result, there’s a lot more squeeze-out, and thus a lot more to clean up, especially since I don’t want any buildup in the outboard end of the tank to interfere with the fit later on.

Anyway, that’s it for tonight. Not a lot to take pictures of, so no photos tonight. Tomorrow I rivet these things in and maybe try and get the fuel sender plates in place. That should give me this week for any touch-up tasks, and to let the sealant cure, and then I can do my water leak test late this week and then finally get these baffles in place!

Whew…I’m finally getting there! I got out to the garage around midday today and started by removing most of the clecos, cleaning out the holes, and then I mixed the sealant and got down to riveting. Things went well for the most part, except that I had some problems with the squeezer sliding around, and I ended up with a fair number of clinched rivets like this: (note: there were some much worse than this, I took this picture mainly because it was the best-lit example)

So after I had Josie come out and help me rivet the forward attach bracket, I took a little break and then went back to inspect the rivets. I ended up marking about fifteen or so that I felt were bad enough to merit replacing, and those got drilled out.  Many of the holes ended up oversize; I suspect this was the problem from the start, since oversize holes can easily promote clinching like this. So I ended up drilling out to #30 and using NAS oversize rivets in most of the drilled-out holes. Tougher to squeeze, but they set nicely and now I can live with the state of the tanks.

Meanwhile my first batch of sealant had cured to the point that it was unusable, so I mixed up a new small batch to use replacing the rivets. That, in turn, meant I had some surplus left over afterwards. This worked out well; I still needed to encapsulate all the shop heads on the nose reinforcement rivets. I had just enough to put a good dab on all of those guys.

One thing I’m debating now is whether to encapsulate the shop heads on the end ribs. These are outside of the tank itself, so they won’t be exposed to fuel, and thus it’s probably OK to skip them. On the other hand, on the off chance any fuel is able to get between the rib and the skin, sealant on the shop heads could be a last line of defense to prevent a leak. I guess I’ll see what other builders have done, and if I’m still on the fence, I’ll go with the old “it’s cheap, don’t be skimpy” line.

So these things are almost done! This week I just need to finish the wiring for the fuel senders and torque the vent line fittings, and I imagine I’ll do a final inspection of the tank internals. Maybe Thursday and Friday evenings I can do my water leak tests, and Saturday I can get rolling on the baffles.

Oh, and here’s one more picture. It doesn’t show anything I built really; it’s actually the big empty shelf space where all these wing and tank ribs used to live. The pile of parts is gradually getting smaller!

Well, that’s it. I think these things are done with the exception of the baffles.

I started out the night by working on the inboard sender plates. I had figured that these might be a bit of a pain, and I was right. The main issue is that the final work on these things has to be done within the confines of a tank bay. First, the wires from the outboard plate and the inboard terminal are pulled into the second bay and cut to length. Next they get stripped and crimped into a ring terminal – this was a lot of fun. Stripping the wires was easy, but getting my crimpers way in there while not pulling the terminal off the wires was quite a challenge. This was especially difficult since the tight quarters made it tough to squeeze the things with sufficient force to get a good crimp. I was glad an extra terminal came with the kit, since I failed in my first crimp attempt.

With those guys crimped, now they had to be attached to the plates. This was even more fun – trying to manipulate a stubby Phillips screwdriver and a combination wrench at arm’s length inside that bay. I found that I could reliably turn the screwdriver about 1/8 of a turn without it slipping off the screw, so needless to say, tightening those things was rather tedious.

With the terminals secure, I mixed up a couple ounces of sealant and put it in a bag for application. More fun here – I needed to coat the exposed terminals in sealant, once again in the tight space. I did the best I could to squirt a liberal amount of sealant on each terminal, then I smeared it all over the place with a gloved finger. I set the bag of sealant aside for later use and mounted the plates to the ribs.

The last step was to get out the multimeter and check for good continuity in the circuit. The general principle of the capacitive senders is that the tank structure acts as one half of the capacitor, and is grounded to the outside of the BNC connector. The sender plates act as the otter half, and are wired to the center terminal of the connector. So the expectation is that the plates have continuity with each other and the center terminal, and no continuity with the tank structure. Similarly, the structure should have continuity with the outside of the connector. All of these checked out (whew!).

One inboard plate, mounted in place:

I also added dabs of sealant along the wire run to hold it to the vent line, and also on the snap bushings to prevent rotation:

I used the rest of the mixed sealant to cover all the shop heads on the end ribs. I did some research on this and the results were unsurprising: Some builders felt it was unnecessary, while others took a “can’t hurt” approach. I agreed with the “can’t hurt” folks, so the heads got coated:

Finally, I unscrewed the B-nut between the vent lines and their bulkhead fittings, coated the fittings with fuel lube, and then reinstalled the B-nuts and torqued them. And that, my friends, completes the tank internal stuff. I’ll probably do a final inspection of everything tomorrow night, just in case anything needs touching up, but I don’t expect any problems.

For the last picture of the night, here’s a family photo:

Now to decide what to do with the rest of the week. Do I take a few evenings off and relax, or start looking at leading edge stuff? I’m very ready to work on stuff that doesn’t involve sealant…

Not much going on today, but I did get the water leak test done on the tanks. I checked the leftover sealant from Monday last night and it was pretty well cured, so I decided to go ahead and to my leak test tonight. Not really much to this, just put a tank in the cradle, plug the drain and fuel feed fittings, and try to ghetto cap the vent line fitting (the leak test kit doesn’t come with that cap; you’re supposed to use a balloon instead).

I have to admit that it felt a little wrong to just start running water from the hose into these things. All the hours I’ve put into them, the electrical connections and so forth, and here I am just blasting it with water. I know everything in here will eventually be submerged in fuel, but still…

I filled each tank to the highest rivet line, then let the water sit for about ten minutes to give any seepage time to happen. And at no point did I see any water seeping out of anywhere. That was satisfying enough to offset the discomfort at getting these things wet.

Right tank filled with water:

And no seeping:

I was hesitant to do this test for several reasons, but in the end, I’m glad I did. It’s a big confidence booster to see these hold water. This still doesn’t mean that fuel won’t seep through somewhere, but it’s a very good sign.

Anyway, the tanks are now sitting upright on the workbench. I won’t touch them until Saturday; that should be enough time for the water to evaporate, especially in the toasty garage. (we’re still getting highs in the 80s here in Houston).

I also spent a little time tonight looking over the Duckworks landing light kit instructions. Doing the cutouts for those will probably be the first thing I attack once I get to working on the leading edges…which won’t be long!

Yup, today was the day, time to close up a tank. I’d kind of hoped to close up both today, but as is par for the course with these things, it took far longer than I expected. I think maybe it’s impossible to accurately estimate the time needed for this stuff, like the sealant somehow bends space-time or something.

Anyway, knowing that a lot of things would be going together, I spent a fair amount of time prepping the work area and getting everything I needed ready to go. To start with, I taped some wax paper over the spar bars on the left wing. This was so that, once the tank was closed, I could immediately set it in place on the spar to be doubly sure that no squeezed-out sealant would give me fitment problems down the road:

Next I put the tank assembly, baffle, and attach bracket on the work bench, and gave all the mating surfaces a good cleaning:

And with that, it was time to mix up some sealant. Lots of sealant, as a matter of fact, because there were lots of mating surfaces to deal with. The essence of the procedure here is to lay a bead of sealant along the rivet line for the skin-baffle joint; when the baffle is dropped into place, it will push the bead ahead and make a nice seal. The mating surfaces of the ribs also get a coat of sealant, plus some sizable blobs at the corners of the tank – apparently these are a very common leak location. Here’s everything all lubed up:

Next, I dropped the gaffe in place. I was surprised how easily it went in…next it gets 100% clecoed to the skin. As I did this, I began to suspect a problem, I was getting almost no sealant squeezing out of the joint as I clecoed, which made me worry about whether I had a good seal – especially coupled with the ease with which the baffle went in, like it didn’t actually do much pushing of sealant. I debated a bit as to whether I wanted to just move forward and hope for the best, but it didn’t take long for me to realize that hope was a pretty bad thing to count on. So I removed all the clecos, pulled the baffle out, mixed up more sealant, and applied a second bead of sealant, further towards the back of the skin. This time, when I dropped the baffle in, I got lots of squeeze-out. (in fact, that whole process of removing and replacing the baffle was pretty hilariously messy)

So with the baffle back in place and clecoed, I went to work on the baffle-rib joints. These get blind rivets, since there’s no access to the inside of the tank, and the rivets get swirled with sealant before insertion to ensure they don’t become a leak path. The attach brackets go on the same way, but here the clearance to the rivet stems was tight, and I ended up grinding a fair amount of material off the nose of one of my pop rivet guns. Nothing like some field-modified tools to help things go together…

Finally, I set the outer baffle-rib joints and attach angles with solid rivets, then finished things off by squeezing all the skin rivets (an operation which left my arms rather tired). This was followed by a rather extensive cleanup session, with most of the time spent cleaning the skin behind the baffles, which will sit on the spar. Again, I don’t want any firmer issues here from sealant build-up.

The baffle, brackets, and such, all finished:

And the completed tank sitting on the wing:

So yeah, this took way longer than I expected, although doing the baffle sealing twice didn’t help much. No way is the right tank going to get closed up on a weeknight this week, so I suppose I’ll leave it for next weekend, because tomorrow we’re going to be out of town for the day. But hey, this tank odyssey is almost at an end!

So yeah, not much done these last few days. We had a Halloween party on Friday, and most of the preceding evenings were spent getting ready for that. Yesterday, in turn was spent recovering from said party. I had originally intended to close up the right tank today, but some events from the past few days made me change my mind.

Something like last Wednesday, I did have time to run out to the garage, install the leak test kit on the left tank, and pressurize the thing. I figured I’d let it sit for a day or two and see if it held air. Sadly, when I went out to check the next night, the balloon on the vent line fitting had deflated. That was a bit of an ego hit, thinking about the possible need to open this thing right back up to fix some leaks. But it occurred to me that the air might be leaking out around the balloon too.

So today I took the tank out in the driveway, pressurized it again, and started spraying soapy water all over the thing to look for air leaks. Lo and behold, the only place where I got some bubbling action was…around the base of the ballon. I made sure to thoroughly wet down the baffle seals, since those were the most likely leak source…no bubbles!

Of course, by the time I did this, it was too late in the day to start closing the right tank, so I guess that’ll wait until next weekend. But hey, the end is in sight!

In other news, I also scheduled my first tailwheel lesson for next Saturday down at Texas Taildraggers. I popped down to AXH this past Saturday to meet Joy and verify that I could, in fact, fit into a Citabria. It’s tight, but it’ll work – the main problem is being able to get full back stick with my gut in the way. Sounds like we’ll probably end up removing the setback cushion and replacing it with something thinner just to ensure I get full control authority. I really should get on losing some of this weight…

Well, that does it…unless I find some leaks in this tank in a week or so, I am done dealing with tank sealant for the foreseeable future. And boy, does it ever give me pleasure to write that sentence.

Anyway, not really a lot to write about today. Mostly it was a replay of closing out the left tank a couple weeks ago. The major differences were me putting the sealant for the baffle on right the first time, and avoiding that whole annoying process of taking the baffle back off and doing it all over again. There was also the flop tube to deal with – I’d safety-wired it to the inboard anti-hangup guide to keep it from flopping around, so I had to be sure and remove that before closing the tank. I also was a little worried about the flop tube weight getting stuck in some excess sealant after I put on the baffle, so I made sure to give the tank a shake after I was done so I could hear it moving around in there.

I also was a little too task-oriented to take a lot of pictures today. Again, just go look at the photos of the left tank if you feel cheated.

Here’s the finished right tank on the spar:

And I got another pic of the workbench. This thing looks like it’s been through the apocalypse or something. I guess it’s thoroughly broken in now…