Sunday, August 29, 2010

Biting the bullet

I had put it off for as long as I could. I have looked at them from every possible angle, agonized over whether or not I had them bent just right in all three dimensions, found other things I could do to further postpone the decision to commit to drilling them, and finally reached a conclusion: it was time to bite the bullet and match drill the longerons. There was nothing to be gained by putting it off any longer.

An excuse I had been using was that I didn't have a suitable piece of scrap to use as a stunt double for the side skins, but I found just such a piece while searching the shop for something unrelated. I also found the missing fuel nipple that gave me such grief. Typically, I did not find what I was actually looking for.

The first step was to get the scrap faux-skin clamped to the longeron.

Then the canopy sill was clamped on such that its edge was overhanging the faux-skin.

In order to have any kind of stability at all for the impending and ultra-critical drilling operation, I had to find a way to hold the assembly in place on the workbench. I clamped it to some of the left over 2x4 scraps from the construction of the workbench.

With the traumatic memories of the 71 hole drilling marathon of the center section still painfully fresh, I dreaded drilling through the longeron material. I needn't have worried. I loaded one of the brand new #30 drill bits that I had purchased from a tent-dwelling huckster when I was at Oshkosh. Truthfully, I didn't see how these bits could come anywhere near the lofty promises that guy was making, but I have to admit that I was wrong to doubt him. Hot knife, meet butter. Those bits tore into that aluminum like a Doberman into a pound of bacon.

Even though the longerons will have to come back off for deburring and priming, I went ahead and clecoed them onto the fuselage just to make sure everything fit.

Once they're cleaned up and primed, I'll be just about ready to rivet on the side skins. Before I do that, though, I'm going to do a thorough audit of any holes that don't yet have rivets in them to make sure that I haven't missed the step that would have filled them. Things are going to start getting a little harder to reach once those side skins go on. 

Wednesday, August 25, 2010

The Cure

It was a much better day at the paying job today. I was able to mostly avoid the business-side professional naysayers, critics, and obstructionists that made yesterday such an ordeal by crouching deep down and hiding in the technical trenches. In other words, I stayed busy developing software and successfully avoided meetings and random interactions with co-workers. Don't get me wrong - I love the people that I work with. It's just in the nature of the job that every time I emerge from my den to head to the coffee area or restroom, I run the same risks (metaphorically, of course) as a gazelle trying to get to the watering hole. Where the gazelle has to be wary of lions, tigers, and bears, I have to be ever vigilant of the co-worker approaching me with a "Hey, would it be hard to..." or "I know you're busy, but..."  Those are invariably indications that I'm going to be pulled into a distraction that I could do without, given my druthers.

Again, don't get me wrong - those kinds of things are the reason I'm there and I'm always ready to lend a hand, but the loss of mental momentum that comes with distractions can end up costing me hours. And every now and then I have to work with someone from outside the organization that doesn't understand our business or why we do the things we do. It's an easy matter for them to throw bureaucratic roadblocks in our way without fully understanding the costs of doing so. That can be frustrating.

That happened yesterday.

Those rare days when I can spend the majority of my day head-down writing code are typically the most rewarding, if not always the least frustrating. Building software applications is a lot like building an airplane. There are moments of hair pulling frustration, there are hours of routine grunt work, and there are bright, shining moments when I suffer one of the Brief Moments of Lucidity(tm) upon which I have built a moderately successful career. Those come at a cost, though. While I'm in one of those spells I tend to write a dozen or so very elegant lines of code that miraculously perform the same function that would normally take hundreds of lines to do.

That's all well and good for the few fleeting moments when I can see the inner workings of the magic with laser-sharp comprehension, but there inevitably comes a day when I have to re-visit that code. Without the magic coursing through my synaptic paths, I sit there staring at it slack-jawed, evidencing the same level of understanding as a shrimp looking at a nuclear submarine.

The fun part about what I do is that every little thing is a puzzle to be solved. The frustrating thing about what I do is that every little thing is a puzzle to be solved. Whether each puzzle is fun or frustrating depends on a number of things, but what doesn't change is that each offers dozens of possible solutions, each having its own set of pros and cons. The beauty of the autonomous nature of my work when it doesn't involve the business-side folks is that I can pick the solution that best fits my mood. I can develop an elegant, artistic solution, or I can choose a brute force approach that throws a little more work to the end users. Today's case was one of those. I could spend hours figuring out how to restructure and import old legacy data into a new database model, or I could just build a screen that would let the end users import the old data a piece at a time as needed, applying the restructuring as they went.

I chose the second option.

It was already mid-evening by the time I got to the hangar to do a little work on the plane, but the weather was perfect for being out there. Just the right temperature, a pretty sky, and a light breeze. Truth be told, I probably should have gone flying. I chose to jump in with some easy RV-12 work instead. The first job on the page was to separate a couple of parts and rivet them to other parts. These little parts had given me some trouble when I was gathering the pieces from the store room to carry out to the remote site. The problem was simple: I couldn't find them! I finally consulted the inventory sheet and found that I couldn't find them there either. That means that they're probably in one of the bags, but I couldn't tell for sure since the bag inventory sheets stay at the hangar. There were also a couple of other larger parts that I couldn't find, but they were further down the page so I decided to worry about them later.  That turned out to be a good decision - I found them out at the hangar.

It was one of those parts that I installed early because it was much easier than it would have been now. I just failed to mark the step in the manual to remind myself. Typical POMOM*

I pulled the little parts out of the bag and riveted them to the bigger parts.

The bigger parts each get riveted to the fuselage using eight of the non-standard LP3-5 rivets.

Oh, remember the mystery holes from yesterday that looked like they should get a rivet installed but the plans didn't say to? Those turned out to be for use with the optional lighting kit. I'm not 100% decided on whether or not I'm going to install the lights, but I'm going to leave the holes open just in case. I mention it now because I no sooner solved that mystery than I found a new one. The posts that get installed with eight rivets have (wait for it....) ten holes.

Those supports will, amongst other things, support the arm rests. Van's left a little busy work on the arm rests; a small area of metal has to be cut out. Just to twist the knife a little, the arm rests are too long to let the band saw get a clean line on the areas that need to be cut out, so some clean up with a small file is required.

As I was deburring the arm rests, I noticed that one of the flanges on the end of each seemed to have been damaged in shipping. It turns out that one should resist the impulse to straighten them out.

Van's, in their typical inimitable style of doing nice things for us right after sticking us with a piece of busy work they could have done for us at the factory, put that bend there so that the part would fit better once installed.

* Problem Of My Own Making

Tuesday, August 24, 2010

Workborne pathogens

Much as with their airborne counterparts, workborne pathogens have a way of infecting low-immunity persons such as myself, often leading to unshakable feelings of malaise, angst, frustration, and outright anger. Two out of four today for me. The latter two, if you must know. Having spent the better part of the day struggling to finish a two day project (today was the sixth day), I was in the mood for working on something more productive and tangible than the computer bits and bytes that I push around all day hoping to form into useful tools,

Lucky for me, I have just the thing for that.

Tonight marked the return to "normal" work. I knew that the tide of struggling with unsatisfying longeron bending had receded when I was directed to countersink some holes in preparation for (you guessed it) nutplates. I'm actually skipping ahead a little bit, which is a polite way of saying that I'm avoiding some work I don't want to do on (what? You guessed it again!) the longerons. Namely, match drilling the holes from the canopy sills into the longerons. I'm putting it off because it requires a .025" thick piece of scrap aluminum. The designers and manual writers at Van's have yet to internalize the fact that the RV-12 generates very little scrap metal (although it did manage to generate quite a bit of scrap fuel line, at least in my case). I'll be cutting some .025 areas out of the side skins soon enough; I'll circle back once I do. Those scraps should be just the right size since the required scrap is to be used to simulate the thickness of the side skins. What could simulate side skin better than, well, side skin?

So, back to my night of relative normalcy. My goal was to get through page 23-03, at the end of which there would be a cross bar and two plates used to support the roll bar mounted to the fuselage. It was the support plates that needed the nutplates. They were slightly trickier than they looked. First of all, four of the holes on each plate are going to take flush blind rivets and need to be countersunk with the special 120 degree bit. The other tricky thing is that the plates get countersunk on the opposite side from that shown in the drawing. It plainly states that fact in the drawing, but if you were suffering from a workborne pathogen infection or something, well, you just might miss that notation. I didn't, but I could certainly see how I could have.

The cross bar clecos into place quite easily.

There are nine rivets in each side, and they're just slightly uncomfortable to get to.

Two things to note here: first, the quality of my photography is going to take a major hit from being out in the less than optimal lighting conditions of the remote factory floor, and two, it took me hand-pulling all nine of those rivets to remember that one huge benefit of being at the remote factory is having access to the air riveter.

Things were cooking along quite well. Too well, as it turned out. Consider this picture, in which it clearly states that there are six LP4-3 rivets to be installed in the rollbar support plate:

Now count the clecos:

Seven, unless I'm still delirious from my recent workborne pathogen infection.

The one in the middle is the one in question. I'm going to have to wade into the fetid depths of the internet to research that issue tomorrow. It was easy and painless to simply ignore it for now.  The rest of the riveting was a breeze, although it might have been better if I hadn't forgotten my reading (and riveting) glasses. One of the problems with working at the remote site is that if I forget to bring something, it's a much bigger hassle to go back and get it.

Fortunately, blind rivets can be read like Braille print.

I'll find out tomorrow if I need to pull another couple of rivets before calling page 23-03 complete. I got pretty close. Close enough, anyway.

Sunday, August 22, 2010

Taking the longeron way around

I'll skip to the punchline: I got the second longeron opened, bent, and twisted. Me being me, though, I took the longeron way around,.And me still being me, I'm going to share every gory detail. It's been awhile since I've worked on the plane, what with various and sundry distractions having conspired to set my goals adrift, but I was determined to get the second longeron completed this weekend. Having the longerons done should, in theory, allow me to get back to the little penny-ante jobs that are so much easier to fit into my daily routine.

Having struggled with the first longeron and thinking that a second pair of hands, and experienced hands at that, might be beneficial when I faced the second longeron, I invited Co-pilot Rick to come by on Saturday afternoon to relive the joy that is longeron bending. I thought I'd get out to the hangar before him to get everything ready to go. Having learned my lesson about how easy it is to bend in the wrong direction, I very carefully and thoroughly marked the aluminum stock to make sure everything would go in the correct direction.

I also went ahead and got the longeron teed up in the vise so we'd be ready to jump right into getting the 5.4 degree opening made on the end.

I had forgotten what a sub-optimal method that is. While it's much better than the Van's suggested "pound it with a sledge hammer" approach (primarily because that method is completely dysfunctional and makes absolutely no sense at all), using the steel pipe coupler has its own issues. For example, there are only two spots on the coupling that make contact with the longeron. Can you see the two "bands" around the circumference of the pipe coupling? What happens in the best case is that the bands create dimples in the edges of the longeron. In the worst, and sadly most common, case only one of the bands works against the longeron. That causes the dimple, but also causes the longeron to bend.

Rick and I talked it over and I decided to follow his suggestion: go to Lowe's and get a longer, smoother length of pipe. He also advocated getting rid of the wooden 2x4 pieces that I was using to protect the longeron material from the sharp edges on the vise pads. We could use the scrap left over from cutting the longerons down to the correct length. Excellent ideas, both. Off to Lowe's we went, which ended up being surprisingly painless. We had no trouble finding a suitable pipe and I took the opportunity to buy a replacement set of lights for the boat trailer. ("What boat trailer?" you ask? Well, follow the 'distractions' and 'adrift' links in the first paragraph - that's what they're there for).  Once back it was just a matter of minutes before the whole contraption was set up and ready to go.

The down side of this approach was immediately apparent: now that I was not only trying to open the whole length at once but was also opening a whole 'nuther chunk of angle. It took even more pressure than it had before to get anything to move. And it had taken a lot of pressure before.

You think that shirt is loud? You should have heard the screeching and snapping of my back!

It looked like we were making good progress, though. Just as I got to the point where it seemed that the JOMT* methodology had gone as far as it could, it seemed to start getting easier. Figuring that I was simply growing measurably stronger in a matter of mere seconds, I did exactly what the JOMT methodology dictates: I gave it another turn. It was at that point that only one of two things could be true: I was either suddenly and shockingly a strong candidate for a spot on the Olympic weightlifting team, or something had gone horrilby awry in the vise. Not surprisingly, the latter was the case.

Huh, it seems that vises are for holding things, not pressing things. Who knew?

Again not surprisingly, we called it quits for the day.

Sunday morning found me at Lowe's again, this time buying a new vise. I was in a bit of a rush and managed to waste a perfectly good chance to pick up a couple of more things for the boat trailer (What? You still haven't looked at those links??) before Lowe's got to its normal Sunday afternoon condition which, in a handful of words, is best described as a pain the rump.  Having recently discovered that I have the upper body strength required to rip the guts out of a 4" vise (yeah, well, whacking it with a 4 lb. sledge hadn't helped, had it?) I went ahead and bought the 6" vise. I thought briefly about buying the 5" but they were the same price, so...

I had to mount the thing to the workbench, but that didn't take long. In no time at all I was ready to start putting the bend in the longeron. Thankfully the former vise had used its last breath on this earth to finish the 5.4 degree opening. As with the first longeron, the bending of the curve is somewhat anti-climatic, assuming that you do it in the right direction. I did, this time.

With that done, I put in the 4 degree twist (made easy with the electronic angle measuring thingy I picked up at Harbor Freight for $23) and called it quits. Well, I did one more thing: I set both of the longerons in their presumptive locations on the fuselage just to see how they looked.

Then it was on to the boat trailer. I figured there was no reason to spend any time troubleshooting the existing lights since I already had a replacement set in hand, so the first order of business was to get the old lights out of the way.

It turns out that troubleshooting wouldn't have taken very long at all.

I was clever enough to not just yank the old wiring out before using it to pull the new wiring through the center pipe of the trailer.

Wiring in the lights was easy enough, but it was kind of like working on the RV-12, too. Put on a light. Notice that I forgot to put on the license plate bracket. Take off the light. Put on the license plate bracket. Put on the light. Put the light on the other side. Realize that there's no way to get the wires to the back of the lights while they're on the trailer. Remove the lights. It goes on like that... you get the picture.

I finally got them installed and went to get the car to hook them up and test them.

Wrong car.

I'd have to go home for the other. I'd have to stop at Lowe's on the way because in my morning haste I had forgotten to buy a new length of rope and a functional quick release clasp for the rope in the trailer winch. It was the broken clasp that had let the boat fall off of the trailer. (Just go read about it, won't you??) The rope was also looking pretty worn out so I figured I'd just replace both.

The trip to Lowe's went just about like you'd expect for a Sunday afternoon visit. There was a long line at checkout waiting behind an old oblivious guy who was so busy chatting on his new-fangled portable telephony device that he was blind to the fact that the cashier was trying to get him to sign the Visa receipt. I bought a $9-ish 50' length of rope and a $2-something clasp and was so surprised to be charged $21 for it that I checked the receipt on the way out. It showed two $9-ish ropes and nothing about the clasp. I turned right around and went back.

"Oh, I don't know why I did that. You'll have to go to Customer Service to get that fixed."

Customer Service had two people working, and both of them were working on the same thing: another old couple trying to figure out how to fill out the application for a Lowe's credit card. They couldn't decide on which name to use. It seemed obvious to me: just use whichever name will stick your estate with the unpaid balance. Duh. I'd like to be able to say I was pleasant and cheerful when I finally received my $7-odd refund.

I can't.

Trading cars turned out to take a little longer than expected too. I still had another project to take care of: we're almost out of bread & butter pickles, so I needed to get another batch made.

With the pickles in the fridge, well... pickling, I ran back to the hangar to test the trailer lights (they worked!) and replace the winch rope and clasp. As I was finishing up, the guy that hangars his Cardinal across from me came over for a visit.

"So, how's your project going?" he asked.

I replied, "Which one, precisely?"

It's been that kind of weekend.

* Just One More Turn

Sunday, August 15, 2010

Crickets in my canister

One could be forgiven for alleging that my belfry is inhabited by bats, what with the heat and humidity today being barely within the upper legal limit for being safe to sustain in a sauna and me crazy enough to be endeavoring to fashion a left side longeron right out there in it. Steeped in it, as it were. Stewing in my juices. One would have to be nuts to work on such a frustration-infused project in conditions such as these. Yes, "he must have bats in his belfry to be out there in this awful muck."

'Tis a fair cop.

But... it's the first chance I've had in days, and it was becoming ever more difficult to suppress the feeling that my project is foundering on the hostile shores of Neglect Island.  I've been busy with various and sundry chores, tasks, obligations, and to be perfectly honest, less energy-sapping forms of recreation. Fearing that all of the precious momentum gathered in my most recent effort for force my will upon the seemingly immutable 90 degree angle of the longeron was dissipating, I felt that I had to at least make an appearance at the hangar today. Having a fresh look at the plans might also serve to build a more complete mental image of the steps involved in finishing my first longeron.

Be careful what you ask for, or so they say. It took only moments to realize that the angle that I had opened on the end of the longeron was complete in that it was the required 95.4 degrees (strikingly similar to the ambient heat index, as it transpires, but those are different types of degrees), but stunningly incomplete in its length. I had successfully opened a two inch length to 95.4 degrees, leaving only another dozen or so inches to go. I had been visualizing this opening as something that was needed only at the far extremity of the longeron; such is not the case. It actually needs to go 12 11/16" inches down the length. So, back in the vise went the longeron, and the steel coupling was called back into play.

You may recall that I had also brought the big mallet into the mix by using it to pound on the handle of the vise to force the jaws of the vice to close on the coupling, thereby spreading open the angle on the longeron. I started doing that again, but soon realized that I needed a more refined and humane method. I also realized that I needed to better protect the vertex edge of the longeron. I put the longeron between two pieces of scrap wood to protect it. As can be seen, there are tremendous forces at work here, and the wood wasn't really up to the task.

In fact, it split apart with a sharp "BANG!" and flew high up into the air, almost reaching the ceiling of the hangar.

I needed a new method of getting a lot of leverage to apply force to the vise. I put the vise handle through the open hole in the end of my big adjustable wrench.

I decided that a thicker piece of scrap wood might be a better buffer and be less likely to get decapitated when the going got tough. Using the wrench, I was able to get a lot more force onto the vise without the brutish and somewhat uncontrollable mallet.

All was going well in getting the angle to open, but there was a little unwanted bending going on as well.

Once I had the full length of angle opened up, I turned the longeron around in the vice and took the unwanted bend back out of it by bending it.

There it is, ready for the next step.

In theory, by having opened the 90 degree angle to 95.4 degrees, I would have imparted an unwanted 2.7 degree angle to both of the sides. No one cares about the vertical side, apparently, but the horizontal (top) side needs to be flush the length of the longeron. To fix that, a "metric crescent wrench" (as called for in the plans) is used to twist the 2.7" out of the top and therefore put the entire 5.4 degrees on the side. Now, there's no such thing as a metric crescent wrench, what with them being infinitely adjustable in both SAE and metric sizes, and any other size you might come up with. So it must be a joke put there by Van's.

Tell me, do I look like I'm in the mood for a joke?

Fortunately, the 2.7 degree bend was very easy to make, even with my Harbor Freight faux-SAE crescent wrench. Yes, that's a joke. I guess I was in the mood after all. Go figure.

It was finally time for The Big Bend. Which reminds me of something that happened at dinner when I was up at Oshkosh. The Jackson Two had come down to where we were staying in West Bend to have a nice dinner. I knew of a place right on the banks of the river where we could sit out on the patio and have the river as a nice background for our meal. The question came up as to just what the name of the river we were sitting by was. I replied that it must be the Milwaukee River. A woman at the next table over interjected with the comment that if it was, in fact, the Milwaukee River, then said river must take a bend that she was unaware of. I replied that I not only guessed that the river did  take a bend, but went a step further and postulated that it very likely took a bend to the west. How ever did I arrive at that preposterous supposition, she wondered. I suggested that maybe, just maybe, that's where the name West Bend came from.

She had never thought of that, or so she said.

But back to The Big Bend. The idea here is that the longeron will provide support for the side skin where it of necessity ends its upward trek at the edge of the canopy. That area of the fuselage has a graceful outward bend to provide room for my bulky shoulders. The way to determine that the longeron is correctly bent is to use the canopy sill part as a template.

It took me awhile to figure out where exactly to position the template on the longeron. The line that I marked 'A' is at the end of the area where I had just finished opening the angle. There is another line marked as the 'START OF TEMPLATE', but the line drawn to show it was open to interpretation. Luckily, a measurement was also provided. If I am reading it correctly (and yes, there is still some doubt in my mind), the edge of the template should line up with a mark 1/4" (or 4/16" if you don't want to simply the resulting fraction arrived at by subtracting 12 7/16 from 12 11/16) from the 'A' mark. The drawing seemed to support that interpretation, so I made the mark and test fit the template.

You can see how much bend is going to be required.

The plans say that the part to be used as the template can be set aside and the opposite part used in its stead (albeit flipped upside down) if the original part doesn't work with the longeron in the vise. I could see that this was going to be the method I would need to use, but I figured that it would have to be done cautiously to avoid making the bend in the wrong direction. Visions of the shipping costs for having to order a replacement longeron sharpened my focus on this issue quite distinctly.  I carefully placed the opposing part, clearly marked with an 'R' to differentiate it from the 'L' piece that will eventually be riveted to this, the left, longeron in place.

The hash marks on the longeron are 1 inch apart. The idea is to clamp the longeron in the vise and apply some bend at each of the 1 inch marks until everything lines up with the template. The plans also warn that a vertical component can be introduced at the time of bending in the horizontal plane, and that one should be on the lookout for that and correct it when it happens. So, in the vise with it, and start bending. The plans suggest checking the bend against the template early and often to ensure that one isn't actually over-bending the longeron. If one is very, very careful, it is possible to match the curvature of the template very precisely. Which it appeared that I had done a simply admirable job of. Still, it's not official until the longeron is matched with the real template, not the inverted stand-in. Which I did.

Only to find that I had very precisely matched the inverse of the bend. Merde! How could I have bent the thing backwards?? Well, simple really. I had neglected to mark the front and rear of the template, and I had managed to use it backwards.

As it turns out, if one is motivated enough, the longeron bending can be very carefully reversed and re-accomplished in the correct direction.  If anything, it turned out even better the second time.

It seemed prudent at this point to forgo the pleasure of doing the right side longeron. I've been disciplining myself into cleaning a new portion of the hangar every time I go out there, so I decided I'd do a little shop-vac cleaning and call it a day. The target area was the back corner of the hangar where the crickets like to gather. My friendly bird hadn't come back, and I could hear a new batch of the irritating buggers back there. The little fiends are clever - they hear me coming when I go hunting for them and quiet down until I give up and leave. They then chirp up (so that's where that expression comes from...) and call me insulting names.

Well, the joke's on them. While evolutionary adaptation has taught them to be quiet when an irate human comes after them, their species has apparently failed to adapt quickly enough to recognize the danger of the shop-vac. It sucked up a good half dozen of them. I'm just crazy enough to laugh out loud when I hear them chirping away inside the belly of the vacuum. In fact, one could be forgiven for saying that I have crickets in my canister!