Wednesday, June 30, 2010

Just flailing around for a few days

Right up until I decided that it would be easier to skip over Section 23 and jump ahead to some of the later sections that require easy access to the bottom of the fuselage, I was quite comfortable in following the enforced linear path of "do this, then do this, when you're done with those, do this." Now that I have options, I seem to have lost my sense of steadfast direction. I could work on the rudder pedals, or I could work on the fuel system, or I could starting painting some of the interior parts, or... well, I just can't seem to decide.

Any decision being better than no decision, I had started down the path of doing the fuel system. That was going fine until I tried to bend the first length of tube. As it turns out, there is a difference between the cheapy tube bender from AutoZone and the more sophisticated tools available from airplane shops. The first bend was to be 90 degrees. And it was! Unfortunately, it was a 90 degree bend with a crimp in the tube for the last 30 degrees. The tube bender didn't have any facility for keeping the tube down in the support channel (I don't know, what do you call it?) and the tube lifted out as I was bending the final 30 degrees. Crimpsville, just like that.  Bummer.

Having ruined an entire four or five inches of precious tubing (which in my prescience I have already ordered more of) I decided that it would be more expedient to buy a proper tool rather than throw more tube into the maw of the crappy bender.  I immediately placed an order with Aircraft Spruce, hoping that the tool would arrive quickly in order to maintain what little momentum I had developed on the fuel system work.

It wasn't to be.

Aircraft Spruce called the next day to tell me that my Visa had been declined. Now I don't want to brag, but overdrawing my Visa account simply doesn't happen. Something else was amiss. I called Spruce and verified that I had entered the numbers and expiration date correctly. We verified my billing address. All of that was correct. On to the bonus questions*:

"Where were you born?"

In a hospital.

"No, what city?"

Oh. Cincinnati, Ohio.

"Is your blood type B-positive?"


"Were you married on May 23, 1992?"

Hold on, let me check my notes since I always get this one wrong.... yes I was.

"Well, it would appear that you're well and truly screwed. What do you want to do?"

Nothing, really. I'd have to say that I'm pretty happy in the marriage.

"No, I meant about the order."


* It should be obvious, but just in case it's not, the bonus question portion of the above is purely fictional. Well, not the part about being flat out of luck with the Visa - it actually was declined.

I went ahead and cancelled the order since  I didn't happen to be carrying my spare Visa card. I then had the CFO contact the kind folks at Visa to see what was going on. The answer? "We stopped approving new charges after a suspicious charge was made on your card."

Really? Suspicious? It was a $1.00 charge at iTunes. In case you're not familiar, let me tell you a little something about iTunes. It is a website that sells songs to be downloaded onto iPods. The cost of the songs? Well, $1.00. So is it truly suspicious to see a purchase of $1.00 from a web site where 99% of the product for sale is priced at $1.00?  As it turns out, yes. It's counterintuitive to me, but Visa says that there is often a "test" charge on a stolen card to see if it works, and charges like they saw at iTunes are very typical ways of testing. Who knew? It seems to me the best test charge would be to actually go out and buy something expensive right away rather than tip off the Visa Police with a measly $1.00 song, but I guess that would come with more risk because you'd be at a physical location where you could be caught. But then again, doesn't the $1.00 test charge carry the same risk, albeit with the addition of a high degree of false confidence? "Yeah, man, I don't know how they caught me. The test charge went fine!"

So now I'm waiting for a tube bender. In the interim I was able to complete the inventory of the wing kit. There was just one little piece missing, and a call to Van's quickly got a replacement on its way to my front porch. They're always pleasant to deal with on these kinds of calls. I always expect an inquisition asking whether I had carefully checked within the folds of the mountain of paper the parts come wrapped in, but they never do. The apologize for the inconvenience, verify my address (and thankfully have no questions about my credit card), and send the part on its way.

I guess I'll start on the rudder pedals now.

But first, a question from the mail bag that I think warrants an answer:
Did you try filling the tube up with fine sand to prevent crimping? Works fantastic for regular copper plumbing pipes.
This is an interesting question because I came across this same suggestion when I was researching tube bending issues. The problem is that I am loath to introduce any more foreign materials than absolutely necessary to the fuel lines. It only takes one grain of sand or other tiny obstruction to block a carburetor jet. It's inevitable that a little aluminum will get in the tubes from the cutting and burnishing, and every effort will be made to clean out the tubes before they are installed on the plane, but I just cringe at the idea of getting a grain of sand in there that steadfastly resists my cleaning efforts but eventually acquiesces to the pernicious demands of flowing gasoline, only to lodge itself in a very inconvenient location further downstream.

That said, if the fancy tube bender also includes more than a modicum of FAIL with its more sophisticated feature set, I will have to look into other means of bending the tubes. I don't think that's going to happen, but if it does I will have to try a more old-fashioned method. What I would try next is tube bending springs. Just like sand, they're dirt cheap (heh! I really crack me up!) and have the benefit of working external to the tubing.

Sunday, June 27, 2010

Just Fueling Around

As with many things, it all has to start with a trip to Lowe's. In this case, it was step one that called for a small amount of pipe thread seal. "Small amount" indicates some arbitrary value greater than zero, and zero is precisely what I had on hand.

It was to be used when installing AN fittings in a couple of the fuel system's mechanical devices. First up was the fuel pump. Nothing fancy about this - I suspect they have them at Walmart.

The mount for the fuel pump was installed in Section 21. While the fuel system is presented as Section 28, I think it ought to be Section 24, coming right after the addition of the nose chin and lower firewall. I chose to do it this way so that I wouldn't have to work around and over the side skins, but it would also have been a little easier to install the pump if it had been done the same time as the mount. I would have bolted the pump to the mount, then riveted the mount to the belly skin. That would have saved me a couple of lacerations caused by rubbing my hands against the edge of the tunnel while reaching down in with a wrench. Another thing that would have helped that would have been to buy the nut drivers I went to Harbor Freight to get. I'm not sure how I managed to forget to do that.

There it is, ready to be the first station on the fuel's journey to the engine.

Aircraft fuel systems always have a valve for stopping the flow of fuel to the engine. The next piece of hardware was this type of on/off valve. In most airplanes, this valve would also act as the selector between two or more fuel tanks. This is the case because most airplanes carry their fuel in the wings. The valve selects which wing tank is to be used, in addition to providing the normal on/off selection. In the RV-12, there is only one fuel tank located back in the baggage area. Having only one tank simplifies the fuel system and also makes it much easier to have removable wings. Not only is fuel heavy, which would make the wings much harder to remove, store, and reinstall, but the added complexity in building a fuel system that could be easily connected and disconnected would be prohibitive.

With that in mind, behold the simplicity of the RV-12 fuel shutoff valve.

The first step is to remove the red plastic sleeve on the handle of the valve and drill a #30 hole in the handle.

The red knob comes with its own instructions, but they are confusingly written. They say to start by drilling a #16 hole and then drill the hole wider to #11. I'm not sure what the #11 hole would be intended for. Drilling it that large would make it much larger than the screw provided to hold the knob on the handle. I drilled the #16 and called it quits.

A couple more AN fittings get screwed in. I used a couple of wrenches to get them in nice and tight. It was blind luck (or good design) that they tightened up right when they were oriented correctly.

With both fittings in, a little box is built around the valve to hold it. 

If you look under the two clecos, you will see a little plate slid under the side wall of the box. That's the way that it fit in the easiest. But...

Darn. The drawing has it on the other side. It would have to be moved. On the correct side, though, the fit was much tighter and it was harder to get the rivet holes completely aligned. Now I've been known to grab a drill and use it to remind some holes just what it means to wear the #30 on this team, but I was reluctant to do that in this case. It took some pushing and prodding, but I finally got it assembled.

The actual installation in the plane was a piece of cake.

A piece of cake which, it would appear, has a cherry on top!

Six little screws

I'm still buggered by those six little screws that have been taunting me since the last page of Section 21. It looked so easy! Just put six screws up through the belly skin and attach them in place with six little lock nuts - how hard could it be?

The first sign of difficulty came when I couldn't reach both the screw head and the nut on the other side of the skin at the same time. That's when I knew I was going to need help. That should have been easy too, but I soon learned that the help available to me within the family was not up to the task. The nuts are pretty tight on the screws and try as she might, young Co-pilot Egg simply couldn't hold the socket wrench tight enough. I enlisted outside help, but it was quickly determined that part of the problem was the insufficiency of my tools., The socket I had was too shallow to allow the inch long screw to be fully threaded onto the nut. The suggested alternative was to go to Harbor Freight on a procurement mission for nut drivers. No problem, that, since I also needed a tube cutter for the pending fuel line work.

Harbor Freight did have nut driver sets, but none of the sets contained the required size. The nuts in question are a relatively exotic 7/32" and the Harbor Freight nut drivers skipped from 3/16" (6/32" for those of you that need a translation) to 1/4" (8/32") with no intermediate stop at 7/32". While the sizing of all Harbor Freight tools is somewhat approximate, I didn't think I could count on a convenient 1/32" spread, so it was off to Sears, Home of Precisely Sized Yet Onerously Expensive Tools. It wasn't all that bad, as it turned out. $3.79 for a deep well 7/32" socket. I wouldn't want to buy a 64 piece set at those prices, but it was okay for a one-off.

With the nut now held firmly in place with the Craftsman Unaffordium socket wielded expertly by a manly helper, there was only one other possible point of failure: the screw head itself. Which is, naturally, where the next healthy application of FAIL came from. The fit of the screws through the nuts is so tight that the screwdriver finds it easier to just eat away at the screw head than to torque the screw into the nut. We managed to get two of the six installed, had to use the Dremel to cut two partially installed but completely ruined screws out of their holes, and ruined the remaining two screws in abortive attempts. I now have replacement screws and nuts on order from Van's, along with a preemptive order for more fuel line because I'm pretty certain how well that job is going to go too.

In the meantime, I placed a request for helpful hints on the Vans Air Force web forum. To date the best reply is "tap the nuts with an 8-32 tap first to loosen them up," followed closely by "use Loctite to keep them fixed in place if you do."

That sounds like a workable plan.

Saturday, June 26, 2010

Finishing the chinny-chin-chin

With just two more pages to go in Section 22, today was the final stretch for the lower front fuselage, or what I have taken to calling "the chin."

 It started out easily enough with the installation of the mount that will hold the fuel flow transducer. That little gadget will measure the amount of gas running through the line and pass that measurement along to the Dynon display. It was simply a matter of installing a pair of nutplates (although I had to counter-sink some holes in the mount despite there being no mention of doing so in the manual) onto the mount, clecoing the mount to the inside belly skin, final drilling the #30 holes for the rivets, and riveting it in.

This is the transducer set in place just to see what it will look like.

At long last, the chin was finally ready to be riveted to the rest of the fuselage. I wanted to have it positioned so that I could get clecos in to hold it in place, so I flipped the rest of the fuselage over. Not only was that somewhat tricky to do alone, it also turned out to be a mistake. No matter how I wriggled and jiggled the chin, I could not get it to mesh with the front of the fuselage.

If turning the big Chunk O' Fuselage over alone was a pain (it was!), turning it back over again would be even worse because I'd have to be very careful to align it with the sawhorses so as to avoid having them place untoward torque on the protruding clecos, thus running the risk of bending the skin. I had to ask co-pilot Egg for assistance. With her help I was able to get it turned over and see what the problem was.

It seems that parts of the chin skin that has to overlap the fuselage skin have notches to allow the flanges to mesh, while other parts do not. The only way I could get the non-notched parts to mesh was to force the fuse skin to slide under the flange of the chin. All of the pushing and prodding had scratched up the skins a little bit, and as this is the bottom of the plane and thus likely to be exposed to relatively wet and dirty conditions now and then, I cleaned up the skins and applied a coat of primer to each.

Here's what it looks like in the area where there's no notch:

Here's an area with a notch:

With the chin finally tentatively held into place with a few clecos, Egg helped me get it flipped back over again for riveting. That accomplished, I release her from servitude.

While I had been having problems getting the chin to fit on the rest of the fuselage, I checked with some other RV-12 blogs to see what the problem was. To no avail, as it turns out, as no one else seems to have had a problem with it. I did see one of those helpful suggestions that I run across now and then that suggested installing a couple of parts from Section 23 before getting too far along with the riveting of the chin. Just as I started the riveting of the chin to the fuselage, I realized that I hadn't installed those parts yet. Doing so now would require either calling Egg back downstairs to help flip the fuselage over again or crawling around on the floor to do the riveting. 

I chose the floor.

With that done, it was just a lot of routine riveting. I thought about hauling the whole thing up to the hangar so I could use the air riveter, but I'd only have to being it right back when done. I just did them all with the manual puller.

The reason I would have had to bring the whole back to the shop rather than just leave it out in the hangar is that I'm going to skip ahead to Section 28. Section 23, which would normally come next, is where the side skins get attached. If I do that now, Section 28, which is the installation of the fuel lines, will be much harder because I'll have to lean down into the fuselage to reach all of the places in the bottom of the fuselage where fuel fittings and lines are to be installed. Better to just do all of that while the access to the areas that stuff has to be installed is much easier, I figure. And better to do it in the temperature controlled basement than in the hot, muggy hangar.

I've got all of the fuel system parts, fittings, and materials laid out and ready to go.

I still have to get those pesky system blocks screwed into place, though, Living in a house full of women, I haven't got anybody that can hold the wrench tight enough or turn the screwdriver while I hold the wrench. My only options at this point are to call for external help or adopt a little boy and wait fifteen years until he's old enough to help.

Tough choice, that.

Kickin' it old school

I've never really understood what that meant. Kicking what, precisely? And what does 'old school' mean? Is it like "way back when," before the focus of our schools was almost completely centered on teaching to the government-mandated competency tests? Well, it's probably another of those phrases that make no sense if one takes them literally. You know, like "temporary tax" or "can we still be friends?" Life has a way of teaching you how to read between the lines when it comes to things like that.

Back in the early 70's when Van's Aircraft was comprised of one guy and one single-seat airplane (the RV-3), building a kit airplane was nothing like it is now. By "kit" the manufacturer typically meant a box of raw materials, a set of plans, and a few pages of notes about how to put it all together. The raw materials were somewhat more advanced than a few boxes of bauxite ore from which you would smelt and shape your own aluminum sheets, but not by much. Even so, the fact that parts like wing ribs and the like were pre-formed was orders of magnitude better than the previous standard: the "plans built" plane. Those truly were old school. You would have to form every single piece from raw stock. By way of contrast, it is relatively rare while building the RV-12 to even have to drill your own holes.

Every now and then, though, we come across a part that we get to "fabricate" on our own. A case in point:

This is one of the situations where we start from raw, off-the-shelf material.

The fact that there were 70 inches of material remaining gave me pause, but for I set the disturbing thought that there are going to be many, many more parts to fabricate aside and concentrated on the two-inch job at hand. It was very straightforward: measure, mark, and cut.

This part is a brace that will help support the weight of the gascolator.

A Gascolator, also known as a main line strainer, sediment bowl or fuel strainer acts primarily as a fuel drain for water and small particles of sediment and is usually found at the lowest point of an aircraft's fuel system. The gascolator is located below the level of the aircraft's carburetor and fuel tanks and on light aircraft is commonly located on the front of the firewall, as low as possible.

Once formed, it needs to be drilled to support a nutplate. This is done by match drilling through holes already drilled into the firewall by Van's. The brace needs to be held in place while drilling through it, and this is apparently a recipe for drilling into your own (or, preferably, a helper's) fingers. As such, the instructions suggest holding the brace in place with a small wooden block. This is exactly the kind of thing "old school" builders would have had to learn on their own. Once learned, though, I have to think that this is a lesson that would only need to be learned once! If not, I would respectfully suggest that such a slow-witted builder might not be the best candidate for building and flying an airplane. A nice set of Adirondack chairs might be a better idea. Padded, preferably.

Once the rivet holes for the nutplate are drilled, the brace can be held in place with clecos. Then it becomes easier to drill the two holes that will be used to rivet the brace to the side of the tunnel.

This raw stock is not alclad, so it has no inherent corrosion protection. It gets primered before final installation.

Do those two rivets look like they're centered on the flange? No. No they don't. That is why I'm not building an RV-3. I like the new school!

Monday, June 21, 2010

[Insert Clever Title Here]

Usually as I'm working on the plane, there's a background process running in my head trying to come up with a clever title for the ensuing post. Nothing really came to me this time. I'm still working on the lower half of the firewall, but I'm very nearly done with it. Unfortunately, I need another little dollop of firewall sealant. It's a shame that I had to throw away about 200 times the amount I am now in need of. This time I'll order one of the little one ounce bottles.

At the end of the last session, I had just riveted on the bottom skin. The next step was to rivet in the cooler stiffener, which was the part that gave me so much consternation as to whether holes that seemed like they ought to be countersunk but hadn't been called out in the plans as having countersinking required truly ought to be countersunk. As a reminder, I threw caution to the wind and countersunk them. The fun with this part was not quite over; it had three different rivet types required to get it installed.

Once that was done, a half dozen nutplates were clecoed onto the inside surface of the bottom skin. That was one of those reach-around jobs where it's a chore just to get the things in place.

On the plus side, the reach-around jobs nearly always end up using blind rivets.

With the latest batch of nutplates finally settled into their final resting place, we moved on to installing Cowl Attach Plates. These are nothing more than flat pieces of aluminum that host a collection of (see if you can guess.....) nutplates. There's a little twist here, though, in that the plates can't be flush with the edge of the skin. They have to be recessed just a little bit up into what will be the engine compartment so that the skin of the cowl will be flush with the skin of the fuselage. This spacing is accomplished with the highly technical expedient of... spacers. 

The spacers are simply thin rectangles of thin aluminum. They are to be cut from "AS3-020," whatever that is. It took awhile to find because I thought that it would be a smallish piece of metal shipped in one of the brown paper bags. There was such a piece of metal, as it turns out, but there was only one rather then the two required. I finally found the strip of material over on the parts shelf.

The fabrication is quite simple: measure out and cut a couple of 1 1/2" pieces and round the corners.

The drawing showing how it all fits together is easy to follow.

The little spacer plates get placed inside the flange of the firewall:

The part that will hold the nutplates only has one hole already drilled in it. You use one cleco through that hole and a clamp on the other side to hold the assembly in place. The parts are flat but the flange has a little curvature, so there's a gap:

I took it all apart again and bent a little curvature into the spacer and the Cowl Attach Plate, then clamped and clecoed it back into place.

Once clamped in place, two more holes are match drilled up through the flange and through the spacer and the Cowl Attach Plate. When it's all still attached would be a good time to go ahead and machine countersink the rivet holes for the nutplates. That will keep you from doing what I did, which is to countersink the wrong side of the attach plates. I thought I was doomed, but it turns out that even though the plates were match drilled in different locations from each other, they were still interchangeable side to side. The holes matched up every bit as well as every other hole I've match drilled, which is to say they were approximately aligned.

That done, the last step on the page called for the use of a Uni-Bit to enlarge the hole the gascolator fuel line (I think) will pass through the firewall to 9/16". I got a kick out of this: "To keep the hole centered draw a cross hair out from the hole for reference." Why did I think that was funny? Because I knew that there was absolutely no way a cross hair was going to keep that hole centered. All it was going to do was graphically display just how off center my hole was!

I started with regular bits and worked my way up in size. I was able to keep the hole pretty centered.

When I got to where I needed the Uni-Bit, I marked the step at which I wanted to stop. Uni-Bits have a way of just keeping on going well after the spot where you needed to stop.

I then proceeded to use the Uni-Bit to get the hole off-center.

I'll burn that bridge when I get to it.

The next step was band saw work and I usually enjoy that. It's kind of like going someplace warm on a winter vacation and enhancing the enjoyment of the temperate weather by comparing it to the awful conditions back home. I like to use the band saw because I remember what a pain the hacksaw was. 

In this case, it was the simple separation of four Support Channels. These will sit on the shelf that attaches the lower firewall to the upper firewall. They will support the rudder pedals, unless I miss my guess. The plans call fro the aft two holes of each to be dimpled, but fail to give any written direction as to which end is the aft. Placing them on the shelf shows the aft holes.

That's not the side that gets dimpled, though. You have to turn it over.

Some matching holes get dimpled in the firewall shelf and, as it turns out, so do some non-matching holes. I got lost with the dimpler and put dimples in four holes that didn't need them. It's not hard to un-dimple, but it was an indication that my head was no longer fully engaged in the process. Rather than press on and risk a much more permanent mistake, I just clecoed everything in where I think it belongs. I'll look at it again when I'm fresher to make sure that I have all of the nutplates in the right places. 

Besides which, I need some fuel sealant to run between the shelf and the lower firewall before I can rivet it together.

Saturday, June 19, 2010

That "Substantial Penalty for Early Withdrawal"


I'm back from the future to tell you something important:

RUN to the phone and call Van's. Place an order for:

FOAM 1/16x1/4 V1110 (4" length, if they ask)

Then go download Page 31-13 from the Archive Pages on Use that page to install the wireway. You won't be able to install the bracket until you do the rudder pedals, but those are coming soon. You will be happy to have this done when you get to page 31-13, trust me.

Don't believe me? Then take a look at this.


You know that substantial penalty for early withdrawal warning you always hear tacked on in the verbal fine print of financial commercials? Well, that's not the only place you'll pay such a penalty.

This morning I finished match drilling and riveting the engine mount brackets to the tunnel walls. The plans have become somewhat cagey with any indication as to which way the rivets should go through the holes. I figure if Van's doesn't care, well, then neither do I. I decide based on the following criteria:

  - Which way is easiest to get at?
  - Which way will hide the wart-like shop head from sight?

If those both happen to coincide, it's an easy decision. Such was the case with the engine mounts. Not only was it easier to rivet through the tunnel sides, it also put the ugly part of the rivets inside the tunnel where they won't be visible.

I then clecoed the tunnel walls, lower firewall, and corner pieces together prior to riveting. 

This was going to be a large enough riveting job to warrant a trip to the hangar. I had to go to Lowe's anyway to pick up a roll of replacement screen (follow that link - it's worth it!) to fix the back porch sliding door anyway, so it was no extra effort to stop by at the hangar on the way. It's something like ten times faster to rivet with the air rivet puller and a hundredth the effort, too, so it was a very simple decision.

What I failed to account for was coffee-speed. Having just finished my morning quart, my brain was firing on all eleven cylinders (most brains only go to ten). That's okay, for the most part, but it can cause problems when attempting to synchronize with the more physical appendages. For example, my hands. My hands, unable to keep up with the blazing speed of my caffeine-addled/enhanced thoughts, tend to either fall behind or rush ahead in hopes of keeping up. This is how I ended up pulling the air riveter away from the rivet in mid-stroke of the internal workings of the rivet puller. I knew something was wrong when I didn't hear the normal "bang" of the rivet mandrel breaking loose and the rivet being set. The gaping hole in the middle of the rivet head didn't look right either. I was sure something had gone horribly awry when the mandrel didn't pop out the front of the rivet puller.

What had happened was that the mandrel didn't break. It actually just pulled through the rivet (without actually setting it - it needed to be drilled out) and jammed in the rivet gun. It wouldn't come out the front because it was still clamped into the inner workings of the rivet gun, and it couldn't come out back either for that reason and because it still had the bead on its end.

I had to disassemble the rivet gun to un-jam it. That wasn't very hard to figure out, although it was something that I was nervous about. You never know if little parts will fall out or a spring will launch itself when the restraining parts have been removed.

It did both.

All told, it took about twenty minutes to get it all put back together and working, Now, you'd think that having learned my lesson about letting my hyperactive brain set the pace, I'd be careful to not let that happen again. As much as I appreciate your confidence in me and my ability to learn a lesson, I have to disappoint you. It wasn't a dozen rivets later before I did the exact same thing.

The last step in the riveting job was to rivet the lower fuselage to the tunnel side flanges. This one was a little tricky. While it appeared that riveting through the firewall and into the tunnel flanges would satisfy both of the decision criteria detailed above, there was the matter of a collection of holes in the front of the firewall that had met whatever arcane criteria is used by Van's to decide whether or not to use their only-when-we-feel-like-it dimpling tool. If those holes needed to be dimpled for flush rivets, why don't the rest of them? But if I riveted from the other direction, they'd be even less flush - the wart heads would be out there for all to see.

I riveted through the firewall. I figure those flush rivets will sit under another part (say, the engine count?) and therefor need to be flush while the surrounding rivets do not.