Joshua's RV-9A Project

Did a couple more odds and ends that needed doing, including some fun contortions to work in the baggage and tailcone areas:

• Installed the DB-9 connector for the autopilot pitch servo
• Installed new nutserts to secure the seat heater relays in their new positions
• Torque seal on the fuel flow transducer bolts
• Installed nutplates for the aux alternator relay and capacitor
• Placed the N-number placard on the panel

Also refit the canopy and forward top skin in order to look at a few things up there:

• Attached wire to and nstalled the strip of LED’s that will be the panel floodlights.  Temporarily connected to the switch to test; still need to run those wires for real and put a connector behind the panel.
• Checked again for clearance between the outer top corner of the EFIS and the panel reinforcement; looks like it should be fine, as planned.
• Checked for catching of the canopy skin when opening; looks ok after filing off a bit more from the forward edge.
• Found where the canopy frame is rubbing on the side seal support angles (which weren’t yet installed when the canopy was previously fit).  Those will need to be filed down for clearance.

A collection of things happening in the shop today.  First thing up, finish the canopy sealing.  The excess Sikaflex was rubbed off the exterior of the top skin and the inside of the rear window and canopy (from gluing the spacer gaps), followed by removal of the rear window spacers and insertion of sealant into those gaps.  Smoothed those over with tongue depressors and all looks good.  Some canopy work remains, of course, but the plexi is in place (which was the big summer goal).

Before the gear can be put on the fuselage, the wings need to be fit, which was skipped earlier in order to get the canopy done in the warmth of summer.  In order to do that , the wings will need to be finished (bottom skins riveted on, which means that they must be wired and plumbed).  Out came bottom wing skins, which needed edge deburring and a few dimples made.  After those were done, the skins were etched for priming along with the rear baggage wall; unfortunately it was too windy to prime until after dark, so that’ll have to wait for another day.

Cleaned up the inside of the wings, which had naturally collected a layer of shop dust and spiderwebs, and set about running the various wires and tubes for the in-wing systems (pitot heat, lights, strobes, AOA, and autopilot).  Everything except the pitot tube was run inside the corrugated conduit which was installed during the wing build.  A hole was drilled in the conduit near the aileron bellcranks, to permit wires to emerge for autopilot and pitot heat; once these wires are in position, some RTV can be used there to prevent any chafing on the conduit edge.  The control module for the heated pitot was installed on the rib near the bellcrank inspection hole, which should allow access to hook up the pitot wires later on.  The pitot tube was secured underneath the aileron bellcrank with an adel clamp to prevent interference or chafing.  All the wiring and tube will terminate at the wing root in connectors, to allow the wings and fuselage to be fully wired in the shop, and easily joined at the airport later on.

Installed the rear window today; not a big deal, after the learning that happened on the forward canopy.  Masked the lines, scuffed, cleaned, primed.  The back of the plexi was primed where it contacts the inside of the top skin, and a thin bead of Sikaflex was laid onto the plexi between the screw holes and the mask line (which marked the skin edge).  The window was lifted and slid into place, and when the screws were installed, there was squeeze-out of sealant around the line, which was the goal; this should keep water from entering via that route. The squeeze-out was tooled down and the tape line pulled, which gave a clean edge; the excess on the outside of the skin will rub off easily tomorrow.

With the rear edge screws in, spacers were inserted between the roll bar and the plexi, and sealant was placed between the spacers.  The spacers were inserted from the front edge, rather than from behind the roll bar, both for ease of removal, and because it allowed for filling in the entire joint on the aft side of the roll bar and tooling it smooth.  This time, the tape was pulled in the tack-free time window (less than 50% humidity today, which helped), and everything turned out alright. While the sealant was out, the spacer gaps on the front canopy were filled in and tooled.

Pillows are a necessity when working inside the fuselage, especially when trying to reach odd angles, or curled up in the baggage compartment.  And laying a 2×4 across the fuselage, spanning the longerons, provides a solid place to sit when working on the canopy or roll bar, and can be stood on as well, which puts the rear bow of the canopy at a good working height when it’s fully open.

After the gluing was done, I spent awhile cleaning up the shop, which had become a disaster area during the canopy work.  What a difference a broom and vacuum make.  Tomorrow I should be able to pull the spacers and fill the gaps on the front, and the canopy will be done, save for the skirts (and sanding the front fairing, but that’ll be later on after it’s had more time to cure hard).

Popped out the the shop after dinner to investigate the progress of the canopy.  The Sika datasheet indicated that cure time for a 4mm bead at 50% RH would be around 24 hours, so I poked at it a bit and determined it looked safe to open the canopy.  No problems there; it has set up quite nicely and appears to have a very firm grip on the plexi.

Unable to resist, I decided to futz with the thing, removing the spacers and cleaning up the Sika joints.  True to what has been said by others, the Sika rubs right off plexi or bare aluminum where there is no primer applied.  The interior paint is a different story, and the Sika appears to have a good grip on that too; this applies to the rear bow where I had to re-smooth the fillet after the skin pulled off.  I’ll rub on it some more after the spacer gaps are filled in and the masking tape removed, but the worst case would be to sand down the smeared areas, mask the canopy, and re-spray the rear bow.

The front cleaned up better than I had hoped; the really awful parts were mostly on the bare plexi, so they peeled and rubbed off.  The remainder actually looks pretty decent — a little bumpy in spots, and not perfect, but I’m much happier with it than I was last night.

As it sits now, the canopy is prepped for filling in the spacer gaps, so I will try to set the rear window and fill these gaps in one session, then fill the spacer gaps in the rear window in one final shot.  I think I should be able to get all that with the remaining tube of sealant.

Today was the day, decided to jump in completely and glue the forward canopy.  A trip to the store was required first, to pick up some 400-grit sandpaper to do the final polishing of the edges, along with more masking tape and electrical tape to set the edges of the primer and sealant.  I also found some long, skinny plastic nozzles that looked like they would work for laying the bead in the tight angles at the front of the canopy.  (Hint: they’re not with the caulk supplies, they were in with the concrete-anchor type stuff, apparently these “mixing nozzles” are made for epoxying anchor bolts into slabs…they come with a spiral mixing thing in them, but it pushes out readily with a nail.)

Before gluing the canopy, I decided to fit the canopy latch lugs to the frame, so that was done, and they were temporarily bolted in place.  I set them so that the latches apply just a slight down-force on the canopy, which will hopefully compress the weatherstrip nicely and seal the gap between the frame and the canopy deck.

On to the gluing…First things first, sand the edges to their final smoothness.  I was going to trim the aft edge of the canopy back, but since all the fitting was done the way it is now, and I’m happy with the fit, I decided to leave the edge where it is, and sand it back once both front and rear plexi are glued in place.  Once the edges were smoothed (which goes faster than I expected, perhaps because I already finished them with 220-grit), I set the canopy in place and laid out the sealant lines with electrical tape: inside sides and back, as well as the outside front on both plexi and aluminum.  Then the canopy was removed for taping the inside front line on the plexi and aluminum.  Both surfaces were roughed up with scotchbrite and cleaned with acetone (on the aluminum; a dry paper towel sufficed for the plexi).  Sika cleaner was applied according to the directions, and allowed to dry.

Once the Sika primer goes down, you have a 2-hour window to set the glue, so a lunch break was taken before starting that part of the job.  The primer is definitely very black, and is supposed to be laid on in a single coat, without any second coating; I had read from others to be wary of brush marks, so I loaded the brush and laid on a good solid coat on both surfaces.  It took a few strokes to get the feel for it, which is why I started with the inside front area…that part doesn’t show through the plexi like some of the other portions.  The primer is left to dry for 20 minutes, and when it dries, it has an almost rubbery appearance and texture.

With the primer set, the canopy was lifted into place and adjusted to match the alignment marks made in 4 places during the earlier fitting (since the rear window was removed to allow entrance and exit to the cabin, the canopy can’t be aligned to that anymore…)  Once things were aligned, 1/8″ spacers cut from vinyl tubing were inserted around the perimeter, and the side skirts clecoed in place to hold the canopy in alignment.  The rear bow was clamped to the roll bar using the 7/8″ spacer blocks made during the fabrication process.  The Sikaflex sealant comes in cartridges for a caulking gun, and the sealant was pumped into the space between plexi and canopy frame from inside; the spacers are left in place for now; later they will be removed and the gaps filled with more sealant.  I had modified a stack of tongue depressors to have the radius I wanted on the sealant, so I used those to tool the joints, and had it looking quite nice.

Next, I climbed out of the fuselage to lay the bead on the front outside of the canopy, which is essentially a ton of sealant squirted out and then shaped into an overly-thick fairing.  Later, once it cures for a couple weeks, this will be sanded down to the needed shape and profile for the front canopy fairing, but for now, I squeegeed it into an approximate shape.

This is where the Sikaflex learning curve comes in.  With all the sealant in place, I cleaned up the tools and my hands, and climbed back into the cabin (that’s an interesting task, through the rear window opening…) to pull the edge tapes.  Doing so, I found to my chagrin that the sealant had already skinned, and the skin was being pulled and stretched along with the tape, the other edge of the skin remaining attached to the canopy frame.  So much for my nicely-tooled joints…I smoothed the skin back down as best I could, though the front inside of the canopy I fear will be rather ugly, since it’s difficult to get good access in that area to re-tool the joint.  Later, reading the documentation again, I discovered what went wrong…the 1-hour tack-free time (which I was barely within) is given at 73 degrees and 50% humidity.  The sealant cures via moisture, and today was upwards of 80% humidity (not to mention over 80 degrees).  Hence, the tack-free time was shortened.  Alas, the only casualty was to the aesthetics of the beads, and my pride.  The next plane (ha!) will be perfect, right?

Removed the canopy from the fuselage for eventual sanding to true up the back edge; I also need to read and see how wide of a gap is needed here to allow proper opening of the canopy.  Fit and drilled the canopy skirts, the countersunk/dimpled, masked off the area at the top which will get sealant, and primed the rest.

Measured out a section of LED strip light to go under the channel that runs behind the seats; this will serve as baggage lighting for loading/unloading in the dark.  The power for this will feed from the center of the strip, so I spooled off some 20ga wire and soldered it on.  While I was soldering, I also wired up the pigtails to the stick grip switches; the pilot’s grip got some expandable sleeving to keep all the wires in a nice bundle, as well.  Both grips’ wires will terminate in a connector at the base of the stick–either a Dsub or CPC for the pilot, and a 3.5mm mini-jack for the copilot, which will be easy to access to allow for stick removal.  I may also have a second PTT somewhere for the right seat, when the stick is removed.

Last but not least, I was able to find a good description of gluing a tip-up (illustrated, even) from Jim Ellis on the Matronics Wiki — I think this is the procedure I will use, minus the screws of course.  Fitting the canopy and then inserting the adhesive should be easier and less messy to handle with only two people, compared to applying the sealant bead and then lowering the canopy onto it.

Off-and-on in the shop today while working through other non-RV projects, eating away at the canopy stuff one bite at a time. The day started by jumping in to dangerous waters: drilling the holes in the rear window.  One hole was drilled at a time, then deburred from the inside, and a cleco inserted–from the inside out, so that the spreading “nibs” on the cleco wouldn’t be putting force on the plexi…rather, the large diameter body of the cleco would be holding the plexi out against the skin.  While it was clecoed, it was marked for the final trim on the forward edge, as well.  Off it came to finish the holes–30 holes, countersinks, and step drill enlargements later, plus final trim of the forward edge, there were still no cracks in the plexi!  It was reattached and held in place with a few screws for fitting of the main canopy.

Marking and trimming of the main canopy was easy enough…by the end of the day, the canopy & rear window trimming was finished.  All that’s left is to do some block sanding on the aft edge of the canopy to even up the line and provide the necessary clearance gap.  A few on-off cycles were required to get everything fitting nicely, including the lower side trims and the little part where the plexi goes under the “ears” of the canopy skin.

Sent an email checking on the shipment status of the SikaFlex supplies.  In other news, my latest order from SteinAir showed up today–the highlights were air vents and map lights, along with the usual wires, connectors, and the like.  I couldn’t resist, and I drilled my overhead panel for the lights, switches, and dimmer knob.  Looks good, I think, and seems like it will work well.  I also put some zip-tie bases up the rear channel of the flap housing, which will hold wires for the flap motor, position sensor, baggage lights, and phone/mic jacks (and perhaps other things, too…).

I decided to put the headset jacks at the aft end of the armrest, in the vertical bulkheads, rather than in the roll bar brace — it seemed like a better place after reading several threads, and thinking about the loading/unloading of baggage, and the possibility of hitting the plugs, or having to remove them each time a suitcase needed to go in or out.  Since the plane will usually have the same people in the seats, the headsets in this position shouldn’t need to be unplugged very often.  It also leaves open the option of having a place (the armrest behind the seat) to hold an ANR box, should one or us end up with a headset that has one…I’ve never liked to dangle boxes from connectors, so this fits with that well.  In any event, I drilled the headset jacks and temporarily fit them in place with the isolation washers.  I’ll need to look at my headset and see what the orientation of the plugs is to get the correct jack position (mic/phone inboard/outboard), for the volume control box to be right-side-up.

After reading a bunch of logs, forum threads, and articles on cutting the canopy, it was time to jump in with both feet.  And it was rather anti-climatic; I think the last “big worry” part of the build is over.  I started cutting before lunch, and by 6pm, I had trimmed the tooling marks off, trimmed the front bend in 3 iterations, and split the canopy in two.  I think I got lucky with the front bend fit; it was very close on the first trim, just a few adjustments required; I wouldn’t have needed to trim the third time, except that I wanted to “creep up” on the right cut amount, and more needed to come off so that it would sit down nearer the roll bar.

After the split, I cut the aft curve of the rear window, and drilled the screw hole pattern into the aft skin.  I could have gone farther, but I need to go shop for some 1/8″ spacers to hold the plexi the correct distance from the roll bar to mark the necessary trims for the rest.  The same spacers will be used to set the bead thickness for the SikaFlex, which should be arriving this week.

All in all, it was nothing like the to-do it’s made out to be…though it did take a couple days to get up the courage to pull out the saw.  I used my quartz heater clamped to the opener track overhead, and flipped the finish kit crate upside down to use for a work stand (my bench is not wide enough to hold the canopy bubble).  With the heater running, it kept the canopy nicely warm for the whole operation.  I did let it sit to pre-heat the plexi for about 90 minutes before starting to cut.  For cutting, I used an oscillating saw gizmo, which worked well; very controllable, even with one hand, though two is better as it does weigh a bit, and less flying debris than the cutoff wheel (though there certainly was debris, it wasn’t thrown across the shop, but pretty confined to the work area.  I used the sanding pad attachment for the same tool to smooth the edges after each cutting pass, with 2 grits of sandpaper.  I can heartily recommend this method/tool, with the caveat that I have not used the cutting wheel to compare.

I also experimented on a piece of scrap with some drill bits and countersinks, to get a feel for drilling the plexi.  Light (very light!) pressure and high speed seems to make the best holes, and I actually had better luck with a regular “used” bit than I did with one that I dulled on the scotchbrite wheel.  After experimentation, I think my plan is to match-drill the holes using light pressure on a #30 bit, then countersink with the 3-flute cutter, again with light pressure, followed by opening up the holes to 5/32″ with a step drill…that made nice clean holes in my testing, now I just have to do it 30 times for real.  (I will be using SikaFlex on the forward canopy and the roll bar, but having the required bead thickness on the aft part of the window would inset the window enough that it would be ugly; I will be using a bead of Sika there as a seal, but the structrual attachment will be per plans, with #6 screws.

Finally, two milestones together on one day…not only the Big Cut, but the project clock rolled to 1000 hours.  I think that’s enough to safely call the project half complete.

A short session this afternoon, in which I was able to get the canopy stops adjusted and file the skins as necessary to allow the canopy frame to open without catching skins.  With the lift struts installed, it’s somehow quite satisfying to open and close it.  With everything playing nice, the forward top skin was clecoed fully down to provide a solid forward structure.

All that done, we lifted the canopy out of the crate and set it atop the fuselage.  And there it sat, while I stared at it for awhile, pondering the why and how of trimming the bubble.  I still have not found a good “guide to cutting” from one who has gone before — there’s plenty written on how to cut or not cut, and the merits of various tools, but precious little on what gets cut off, the sequence of trimming, and the like.  More research ensues…temps in the low 70′s today, maybe it will warm up at the end of the week.

The night that the last few weeks have been building up to (no pun intended…) finally arrived.  The primer and paint were dry, and canopy frame is now riveted. Not a bad job overall; before riveting, I had to back-drill the holes for the screws that hold the lift strut blocks through the skin, and dimple.  The frame skeleton was riveted first, then the spacer strips attached (there are 5 keeper rivets per side).  The aft bow was riveted together, after checking for fit against the roll bar and clamping to the table to hold it in position.  Then the bow and the frame were joined, and the top skin added.  For ease of riveting, I lifted up the aft end of the frame by suspending it from the garage door track.  Once everything was together, I riveted the aft edge of the stiffeners to the skin, then clamped the frame flat to the table to cleco and rivet the forward edge of the stiffeners to the weldment.

True to the tales, it shifted a little during riveting — the aft width is nearly perfect, but the front sides near the splice plates had narrowed a bit, and the skin seems to have shifted forward, as it catches the forward top skin now.  A couple strategic pulls took out most of the side gap, and the forward skin can be filed down a touch.  The numerous on-off cycles have broken in the hinge blocks, so the hinge goosenecks slide in easily now.  I won’t worry about perfecting the fit until the next time, when I rivet the forward stop brackets and fit the struts, so I can correct for any flex added by the struts at the same time.

Punch list before fitting the bubble:

• File forward edge of canopy skin
• Fabricate and install (6) straps on stiffeners
• Install stop brackets and lift struts
• Drill hole and install bushing for wiring passage to roll bar brace…install tie-wrap bases on vertical cable path
• Run extra coax aft for GPS antenna?
• Rivet aft top skins
• Read up on fitting process and where/what to cut

Allison has discovered the joy of paint schemes, so I gave her the JetGlo color chip booklet…a side trip yesterday to a beach just a few dozen yards’ walk from an airport (KDYT [airnav]) was good inspiration, I think.  Three hours in the car, just an hour by plane.