Fuel Tanks

So the main order of business tonight was to seal over the shop heads of the stiffeners I riveted Friday night. But since applying sealant isn’t an exact science and I hate to waste more of the stuff than I have to, I’m trying to keep a few small jobs prepped so I can make use of leftover sealant. Tonight I decided my “extra” items would be some of the connections for the capacitive tank senders. This whole system won’t go together completely until the tanks are almost ready to be closed up, but some reassembly can be done. One part is the wire connecting to the BNC connector on the inboard rib, the other is the wire to the outboard plate. Final assembly will consist of attaching the unterminated ends of both these wires to the inboard plate.

So I dug out the soldering iron, which I don’t think has seen the light of day in quite some time (actually, it took me half an hour to find the thing). The BNC connectors get a short ~12” wire run; it only has to cross one bay in the tank. Soldering those took far longer than I’d like to admit. Yeah, it’s been a while, I’m kind of rusty. The second wire run was ~48”, since it has to cross four bays, and this run terminates with a simple ring terminal. The instructions recommended soldering in addition to crimping here for additional security, which I did.

With the soldering done, I scuffed and cleaned the mating surfaces on the inboard ribs and the outboard sender plates, along with the terminal assemblies, then put everything in place, the BNC connectors in the ribs and the ring terminals screwed to the plates. Then it was time to mix up the sealant.

After putting the sealant in my application bag and dabbing all the shop heads, I had plenty left over for all the terminals, so I went to town. These actually use a fair amount of sealant, since you want to thoroughly cover any exposed wire as well as the terminals themselves. I ended up sculpting little volcanoes on the BNC connectors, and the ring terminals – well, I’m not sure what to call those other than straight-up blobs.

In other news, I got my final shipment from Spruce today, which contained my Whelen landing lights:

I’d previously bought a little 12V power supply specifically for testing stuff like this, so of course I went out in the driveway after dark to test them out. They’re pretty bright, though in fairness I’m barely 20’ or so from the fence. I should probably be testing them individually to get a better real-world look; in practice, one light will be aimed as a landing light and the other as a taxi light. At least that’s the plan for now…

So I kind of took most of the evenings this week off the project. Tuesday I had a friend from out of town come by for a bit, and Thursday I did an EAA webinar on recreational aerobatics. I could have worked the other two nights, but really, my next sealing task would be the main ribs for a tank. I knew that would be a time-consuming job and felt it would be best saved for the weekend.

Of course, this morning, I abruptly remembered that I had a few more prep tasks to do. The ribs for one tank still needed to be scuffed, and I hadn’t yet dimpled any of the ribs. So those were my first two tasks today, hitting the edges of the left tank ribs with the scouring wheel, then dimpling away with the squeezer. I wish I had thought of the dimpling earlier, specifically before I installed the BNC connectors for the senders. They got in the way of dimpling a couple holes. But the forward most holes were also inaccessible with the squeezer, so I had to get out the tight fit dimpling tool anyway.

Then it was time to clean all the mating surfaces, mix up a big batch of sealant, and go to town. Here I am painstakingly paying a bead of sealant on a rib:

I had a little moral support along the way:

Then, once I’d gotten each rib started, it was time to cleco away. I was slightly concerned that I might come up short on clecos and need to pull some from the main wing skins, but it turned out I had plenty:

Before going in and working the squeezed-out sealant into fillets, I took care of some small tasks with my leftover sealant. Item #1 was the tooling holes in both outboard ribs. These holes were covered by the access plate reinforcements in the inboard side, but out here I had to do something different. Some guys buy 3/16” rivets just to squeeze in these holes, but I chose to just use some 3/16” pop rivets I had lying around. This isn’t structural at all, so no concern with strength. And both rivets got covered with sealant on both sides anyway:

I also put the nose reinforcement plates in place and clecoed them; I’ll rivet them tomorrow:

Then it was back to the tank, where I spent some time reaching in and digging around with a popsicle stick to make the seams nice. Most everything looks good, though there are a few spots where I’ll probably add some extra sealant later on:

Finally, before calling it a night, I decided that I wanted to clamp the trailing edges of the skin behind the ribs. I figured this would help pull everything into assembly and maybe get a little extra squeeze out from the back ends of the ribs. Problem was, I only had two clamps big enough to cover the width of that area, but I needed to distribute the force all along the length of the skin. Oh, I know, I can use the steel pipe lengths that will eventually be the aileron counterweights:

And that’s it for now. This is the easy part; tomorrow we get to go in and buck a whole bunch of rivets. Later on I’ll fay seal the end ribs in place, but first I’ll have to put the sender plates and the vent tube in. Which reminds me, I still need to install the vent tube fittings in the inboard ribs…

Well, I suppose I was overdue for a good frustrating day like this. I don’t think I’ve had one of these in a while.

Things seemed to start off OK, save for me noticing that I’d neglected to dimple one of the end ribs. Didn’t really affect today’s work, but I went ahead and fixed it up. From there I set about doing some general cleanup in preparation for shooting rivets. Most notably, I pulled 2/3 of the clecos, figuring on riveting those holes then coming back for the last set. That led to the fun of cleaning out the holes. I’m starting to become a little disenchanted with this whole fay sealing process. It’s supposed to significantly cut down on the mess of riveting, and I suppose maybe it does, but the big downside I’ve seen is that when I come back the next day to rivet, I find that many of the rivet holes have sealant buildup that needs to be cleaned out prior to riveting. Bottom line, I’m not sure if I’ll continue using this method for the rest of the project or not.

Anyway, with that done, I mixed up a batch of sealant and went to work dabbing the holes, then inserting the rivets, and finally going to town with the rivet gun. This was going fairly well – actually, I was surprised how easy it was to rivet these solo. I’d kind of been expecting to need more help here. That is, until I got distracted and dropped the bucking bar, which made a couple of nice little dents in my skin:

I suspect I can probably knock those back down with a flush set and a bucking bar on the inside, though the inner curve may present a problem. So yeah, from that point on I was much more careful to keep a grip on the bar. After a while (and a lot of pairs of gloves), I’d set all the rivets that were accessible with the tank in the cradle. That left the toughest ones, all the way up at the nose of the tank.

I was working on setting those when I ran across another problem: the forward most skin hole on one of the ribs had a badly-done dimple. I guess when we were dimpling this hole only got a soft hit and we didn’t catch it. Kind of a problem now, since everything is more-or-less permanently assembled. I tried dimpling with the pop rivet set I have, but if it did anything I couldn’t tell. Not really surprising though since the void between the skin and rib is undoubtedly filled with partially-cured sealant. I’m thinking maybe I can put a NAS1097 rivet in this hole; they need a much smaller dimple, and while that compromises strength a bit, it’s probably better than making a countersink. I may shoot an email off to Van’s for some support here just to be sure.

And this led to the final nail in the day’s coffin. I’d gotten Josie to come out and help shoot those nose rivets; they were too far in for me to rivet solo, even with my gorilla arms. Previously, I’d noticed the rivet gun sporadically making a buzzing sound instead of hitting when I squeezed the trigger, but it was rare enough that I didn’t think much of it and just kept pushing through. But once Josie came out and we’d shot a few rivets, the thing wouldn’t shoot at all, just buzz. I pulled it apart and I’m pretty sure I found the problem:

All told, this is better news than I was prepared for. I was about ready to just order a new rivet gun come Monday morning and replace this beat-up US Tools thing. But if I can replace a ten-cent O-ring and get back in business, that’ll be much better.

So in summary, I got about 80% of the rivets set, and the rest will have to wait until I get the gun fixed. Normally it would bug me to wait this long between fay sealing and riveting, but given how tightly the tank skins fit, and that the remaining rivets are at the nose where things are the tightest, I think there won’t be any issues here. So I’ll leave you with a photo of the mostly-done left tank:

Coda: I didn’t really get at it earlier, but today I kind of hit a point where I was really tired of messing with sealant. I’ll be glad when I’m done with this stuff. On the other hand, this rib riveting is probably the most challenging part of the tank build, so at least I’m about halfway through that.

Well, the general plan for tonight was to try and finish riveting the left tank, but first there were some prep items to tend to. Most importantly, I needed to deal with that mis-dimpled hole. Someone suggested using a NAS1097 rivet in the hole, but when I tried that, it sat a little below flush and I didn’t really like the look. So instead, I decided to countersink the hole. I knew the skin was thick enough to countersink – after all, the baffle holes in the rear of the skin get countersunk instead of riveted. And in this case, with a slight dimple in place, I’d be taking even less skin off. There will be a slight void between the rib and skin, but it’ll be filled with sealant and shouldn’t be in a high-stress area.

I’d also noticed on Sunday that a few rivets seemed to be sitting high, so I went back and inspected all the rivets I’d set. I found about eight or so that I felt were high enough to necessitate replacement. I was a little worried about this – in my mind, the rivets would be welded in place with sealant and impossible to remove. But in truth, it was no trouble at all. As a side note, I’ve been doing so much structural stuff lately that I’ve forgotten how easy it is to drill out these skin rivets. I’d just broken out a new drill bit before getting started, and I actually drilled all the way through one rivet instead of just going partway into the head. Fortunately, I was right on center and didn’t damage the hole.

One of the proud rivets before drilling out:

Next, I cleaned the dimples for these holes and used a long 426 rivet to test fit for flushness. I ended up spinning a countersink bit in each of the holes to help remove leftover sealant and to get a better fit. Finally, I gave each dimple a good scrubbing with an acetone-soaked rag. At this point, though, I decide that it was too late to go and mix up sealant and rivet. Better to do that when I have plenty of time ahead, even if I’m only shooting a few rivets.

So instead I figured I’d just work on some minor tasks. First up, I got out my outboard ribs, to which I’d sealed the reinforcement plates earlier. After cleaning the surfaces up a bit, I riveted those in place:

Next, I needed to modify the snap bushings to accommodate the fuel sender wires. These bushings are intended to route the vent line inside the tank, but since I’m using the capacitive senders, I’ll need to route a wire through them as well. This was easily accomplished with a cutting bit in the Dremel, though it made a mess of plastic shavings:

Quick look at a modified bushing with a piece of sample tubing and wire routed though it:

Finally, I wanted to finish up the hinge pin setup for the trapdoor in the right tank. Earlier, I bought an extra piece of hinge pin stock from Spruce just for things like this. I cut a piece to the length I wanted and put a couple of right-angle bends in it. The idea was to drill a couple of tiny holes in the rib and use safety wire to hold the pin in place. Here’s an initial test-fitting of the setup:

I’m not all that pleased with my wiring job, but that’s OK, cause I’ll probably just remove this again for the time being. Having the trapdoor flapping around would be a pain while assembling the tank, so I’ll reassemble it before the baffle goes in, and probably add some sealant on the works just to help ensure everything stays in place.

So maybe tomorrow I can finish this riveting. I also still need to put the vent line fittings in place and get them sealed…

Whew. I think I’m starting to see the light at the of the Proseal tunnel…

I still haven’t gone back and finished riveting the left tank, both the nose rivets and the ones I drilled out. We’ve talked about doing them a couple night this past week, but always either Josie was working late, or I was feeling tired, so forth and so on. I wanted to do the right tank ribs between tonight and tomorrow; Friday night to fay seal the ribs and Saturday to rivet them. I’m trying to not let a lot of time pass between those two tasks, on the off chance that sealant buildup there might affect the fit of the tank. Anyway, my conclusion was to go ahead and do all the solo stuff possible on the right tank, and then we can just knock out the two-person rivets in one session instead of two.

I was mindful tonight of all those proud rivets on the left tank, so I started by inspecting all the dimpled holes. I went so far as to insert a test rivet into each one to ensure that it sat satisfactorily flush. Almost all the holes were fine, but I did touch up a couple with the countersink for a good fit. It was a little tedious, but better than dealing with unexpected stuff.

Then it was into familiar territory of cleaning the mating surfaces, mixing up the sealant, painstakingly applying it to the rib flanges, and clecoing everything. I took a picture before I got the tank all clecoed together to kind of show the process:

I start by putting two clecos in the upper nose of each rib (shown above), just to hold it in place while I take care of the other ribs. Next I add a few more clecos on the top before putting the whole assembly in the cradle. Then I put two clecos in the bottom rear of each rib, which pulls the skin tight all around, and then it’s just a matter of clecoing all the other holes. Then I get to go bad with a popsicle stick and work the squeeze-out into nice fillets. Finally, as before, I used the aileron counterweights and clamps to pull the trailing edges of the skin together:

Since I had a little sealant left over, I also went ahead and installed the vent line fittings in the inboard end ribs:

That second picture is the inboard side of the rib. I probably used way more sealant than necessary, but I figured that covering the entire nut and washer assembly was a good way to ensure no leaks. There a couple of maxims I’m keeping in mind from former builders: the first goes “Proseal, is cheap, so don’t skimp.” The second goes “If you look at a joint and say ‘I don’t think that will leak,” it probably will. If you look at a joint and say ‘There’s no way that will leak,’ you’ll probably be fine.”

So that’s it for tonight; tomorrow morning will be messy time again, doing all the riveting. Hopefully I won’t drop the bucking bar this time. Or maybe I should, so the tanks will be symmetrical…

I probably shouldn’t get ahead of myself here, but I might be to the point of closing these things out next weekend! I will be really, really, really glad to be done with this stuff.

Yay for proseal! Today wasn’t so bad working with the stuff, I had more of an attitude of “I’m getting closer to being done with this” instead of “ARGH THIS IS ANNOYING.” Which I suppose is a good thing.

Anyway, I continued with my theme of being much more thorough about dimple preparation. In addition to my usual routine of getting accumulated sealant out of the holes, I hit each dimple with a countersink bit to ensure there was no buildup of partially-cured stuff. This added a bit of time to the prep, but it beats drilling out bad rivets. Other than that, it was the usual routine of dabbing sealant in each hole before setting the rivets. At least I didn’t drop the bucking bar this time.

On the other hand, when I first tried the rivet gun, it did its buzzing thing again, and for a moment I thought the entire afternoon was going to be a bust. But it worked after that and I had no problems. Still, I’m kind of wondering if I want to keep trusting this thing. I bought it secondhand and it already looked old and beat up, so maybe a new gun would be a worthwhile investment after all.

Also, after finishing up the riveting for the day, I used the leftover sealant to seal the noses of the ribs in the right tank. Previously I hadn’t been sealing this area, my rationale being that there are no rivet holes there and thus no need to seal. But after doing some research on VAF, I found that many builders felt this was a potential leak path where fuel could maybe get under the rib. In the end, I fell back to the “sealant is cheap, don’t skimp” truism and lathered that area up. I’ll do the same on the left tank at some point too.

Rib noses all sealed up:

Hopefully tomorrow we can get the rest of the nose rivets done, the ones I can’t shoot solo. I also need to replace the rivets in the left tank that I drilled out. And now that I know that drilling out sealed rivets isn’t a big deal, I may go over the left tank again and be more liberal about proud rivets I want to replace. Finally, I need to re-rivet the inboard tank attach nutplates to the right spar…that’ll be a two-person job too.

I’m still hoping I can be ready to seal the baffles next weekend, though that may be asking a bit much. Maybe if I can really commit to being productive int he evenings this week…

So yeah, once again, we didn’t get around to actually finishing up the tank riveting today. That was the original plan, but when I went out to remove the last of the clecos from the right tank and clean out the holes, I ran into a problem. I’d estimated that my gallon jug of acetone was about 20% full, but boy, was I wrong. It had about a cup or so left in it. Normally, this wouldn’t be a huge issue – just run to the store and get a new one – but we had a busy day, and we were already sort of cramming in a couple hours for the skin riveting. Throwing in a 20-minute trip to Home Depot seemed like enough to make it smarter to just put it off, so that’s what I did.

Instead, I focused on getting everything as ready as possible for when we finally do do the riveting. I went ahead and cleaned all the remaining holes in the right tank, cleaned up the dimples with a countersink bit, and test-fit rivets in each hole. Next I did a secondary inspection of the left tank – when I went after the proud skin rivets before, I was picky about which ones I’d drill out, since I expected it to be an ordeal, Now that I knew it was much easier than expected, I went back and was more liberal about marking rivets I didn’t like, and I ended up drilling out about ten more. Then I cleaned those dimples with the countersink as well, and test-fit rivets to make sure everything looked good. With that, I set the tanks aside; tomorrow I’ll just need to give them a final acetone cleaning before we start setting them.

The only other thing I wanted to get done today was to rivet in the inboard tank attach bracket nutplates to the left wing spar. I found when doing these on the right wing that back riveting seemed to be the most reliable method, so I did that again here, though the challenge was increased by having to work in the narrow space between wing walk ribs with the rivet gun. Instead of a back rivet plate, I used a tungsten bucking bar, which I had Josie hold in place:

The technique worked fairly well, although one of the rivets sat a little proud. And this was when I made a stupid error. I wanted to try and knock the head down a bit, so I tried hitting it with a flush set – but it was too close to the spar reinforcement bar for me to really get on it. So then I had the “bright” idea of using the back rivet set on the thing. It’s a smaller diameter, with a flush face, much easier to get in tight by the spar bar.

Except I did this:

Yep, a nice little ding in the spar web. Some VAF research indicates that it’s most likely OK to dress the nick out and alodine the newly exposed aluminum, so I think I’m OK here. Still, given that this is the main wing spar, and particularly the inboard portion which will see some of the highest loads, I’ve already sent off an email to Van’s builder support. This is not an area where I’m willing to take the “eh, that looks OK” approach. Again, I’m 99% sure this is OK, but I definitely want a second set of eyes on this.

So here we have yet another recurring lesson: It’s a good idea to take some time and think about these things before just shooting from the hip. That’s when stupid things happen.

So once again, riveting didn’t happen tonight, cause Josie had to work late. However, I got a prompt response back from Van’s about the ding I put in the spar, and as I expected, the word was that it was fine to remove the ding and prime the area. So that was what I started on tonight. Fairly simple repair, though being right up against the spar reinforcement bar made it a little tight at times.

The original ding:

First I applied a few layers of painter’s tape to the spar bar to prevent accidentally damaging it:

I started with a flap wheel in the Dremel at low speed, and that made short work of removing most of the ding. But the spar bars prevented me from getting the entire thing with the flap wheel, so I ended up using some 220 emery cloth wrapped around a bucking bar to sand out the rest of the ding:

Next I smoothed and blended the area first with a scotchbrite pace, followed by 600 grit sandpaper:

Then, after a good cleaning with an acetone-soaked rag, I masked the area for priming:

And here we are, good as new. Sort of: (I need to replace that other rivet though, it’s really sitting proud)

After this, I did some general cleanup, and in the course of that, I noticed an amusingly dumb oversight: on one of the left wing Z-brackets, I only riveted one side of the nutplates. Not quite sure what happened that night…

So I pulled that bracket off and squeezed those rivets. Then, since I already had the tools out, I went ahead and unbolted al the other Z-brackets and deburred the rivet holes to the tanks. One more little task out of the way before the tanks get closed.

So yeah, that whole notion of maybe closing the tanks this weekend didn’t work out. It probably would have been a stretch anyway, but Josie had a really crazy week at work and worked later every night (sometimes into the early morning hours). So she wasn’t able to help me rivet at all during the week.

Today, however, we finally went to town and finished the skin riveting, including replacing all the rivets I drilled out of the left tank. Once that was done, I used the leftover mixed sealant to seal up the nose rib regions in the left tank, and also to get the tank drains in place. I didn’t bother with the fay sealing stuff for those; while it was messier, there were only six rivets each, so no big deal. Before putting the drains in place, I pushed a wad of masking tape inside the drain hole. I’d read about guys using Q-tips to clean sealant off the drain threads, but that sounded like no fun at all, so I decided to make it my mission not to get sealant on the threads at all.

Drains with the holes plugged:

And after riveting:

After taking a quick break inside, I came back out and mixed up another batch of sealant to cover all the shop heads with. This batch also finished off my first can of sealant. According to Van’s, it’s possible to seal both tanks with just one can if you’re frugal, but it seems like a stretch to me. I’ve wasted some sealant to be sure, but not enough to do the end ribs and baffles.

Anyway, I put that batch into a bag and went to town on the shop heads (though it occurs to me as I write this that I neglected to get the shop heads on the drains):

Finally, as an aside, I actually had the workbench cleaned off at one point today, which allowed me to reflect on how well-used it looks now that it’s been repetitively stained during the tank-building process:

I also worked on the flop tube. I’d been thinking all along what a pain this was going to be; the tube needs to be safety-wired to the bulkhead fitting, and I’d been imagining all along that I’d have to work inside the tank to do that. It abruptly occurred to be today that I could just put the two pieces together outside the tank and do the safety wiring with plenty of room…so I did it:

That’s it for tonight. I started to go ahead and put the vent lines and other interior stuff in place, but I want to go back tomorrow once the sealant has cured a bit and mash down all the blobs on the shop heads. The vent lines would definitely get in the way of that, so better just to wait. Tomorrow I should be able to get all the interior work done and put the end ribs in place. Next weekend really should be the time to close these things out. I imagine I’ll go ahead and do the access covers to, then let the tanks sit around and cure for a couple of weeks before I do my leak tests.

Well, lots of stuff to do around the house today, but I did manage to get out and put in a couple hours of work. I did a thorough inspection of the sealant I put around the shop heads last night, and in several cases massaged the blob around with my finger to get better coverage. About ten or so rivets needed extra attention (i.e. more sealant), so I marked their location as I went. Then I mixed up a small batch of sealant and touched up those rivets. There were also come locations between the rib flanges and the skin where I didn’t feel I had good sealant coverage, so I touched up several of those as well. I really don’t want these things to leak.

This should free me up to work on the interior stuff after work this week – which really isn’t that much, just routing the vent lines and installing the fuel sender plates. I should be able to get the end ribs in this week, and then this weekend really should be the time that I close these out.

Not much to take pictures of today, but here are both tanks on the workbench: