(page 31B) A mechanical engineer doing wiring -- what could go wrong?

The wires in my hand are just for effect.  They actually came out of my antique car when I foolishly decided that the warmed-up LT1 engine I had put in it didn't have near enough horsepower.  With the

money I proceeded to spend on the followup engine I could have easily bought the firewall-forward kit for the airplane.  We live and we learn (I hope).  Money is shaping up to be a major player in how soon I can "slip the surly bonds of earth."  The avionics and engine, which is what I will have left to buy after completion of the "finish" kit, represent more than half the total cost of the airplane.

Given that the airplane will have sophisticated avionics, autopilot, etc., I've always contemplated doing the wiring with a sense of dread.  So far, the dread has been justified.  I immediately discovered that some of the holes shown in the pictures in the build manual through which the wire bundles must pass were not present in my kit, owing, I suppose, to the slow pace of my build.  During the time

lapse from when I order the sub-kit to when I actually get around to working on it, the kit has evolved.  No mention of this is made in the build manual, so step one was to drill the missing 3/4-inch holes in the instrument shelf and three fuselage bulkheads.  The left arrow in the picture shows an array of cooling holes that also aren't there, but I'll put off drilling those until I figure out what goes there.

The step drill shown worked well for these holes.  BNC connectors and two fat wire bundles go through these, so the required snap bushings will be slit, placed over the wires, and snapped into the holes later.

The cooling fans were installed at this point.  After hooking them up to 12 volts, I discovered that they blow in opposite directions relative to the fan housings, and one runs at a higher RPM than the other.

The Mothership said it's OK.  Simply flip one over to get the correct direction and the slow one will probably come up to speed after it "loosens up."  It would be so easy to replace one now rather than later, but I'll trust Van's.

The wire bundles and the three coax cables go over the rudder tube.  A wire tray will be added later. The bolts holding the rudder tubes up must be loosened in order to allow the BNC connectors to pass above the tube.  They still have to be forced through the gap.









In the picture at right, both wire bundles have been run.  Various wires are identified and pulled out of the bundles as the bundles make their way aft through the various snap bushings, with some wires going all the way back to the tail cone area, others stopping at various locations.

The build manual stresses repeatedly to run each wire individually from the instrument shelf to its final destination.  I didn't do it this way.  Instead, I laid out each bundle on the shop floor and taped the end of each shorter wire to the main bundle, then ran the whole thing through the various Adel clamps and snap bushings.  The sticks have been installed in order to run the wires for the push-to-talk switches.  Issues with the sticks will be addressed in a separate post.

Much time was wasted removing snap bushings which the build manual had me install in various bulkheads and seat ribs.  With the bushings in place, various wires with their already-installed connecters won't come close to fitting through the bushings, requiring that the bushings be removed and the wires run through the bare holes.  The slit bushings can then be placed over the wires and finally snapped into the holes.  The problem is that the already-installed bushings are difficult to reach and remove in the tight spaces.  If I were building another airplane I wouldn't install any of the bushings in the fuselage until all wires are run.  The build manual warns of this, but way too late.

Most of the builders who have gone before me agree that delaying the attachment of the tail cone is the smart thing to do (also not suggested in the build manual).  This allows easy access to the rear

bulkhead and also makes it easy to turn the fuselage on its side. Having the fuselage on its side allows me to comfortably sit while routing the wires.

One of my students dropped by and offered to help with the fine adjustments.

Service Bulletin 12-11-09 Part 2: Landing Gear Beef-up

After the RV-12 fleet started to accumulate a significant number of hours, several suffered wrinkled side skin following hard landings.  Close inspection revealed cracks propagating from the bolt holes in the  center-section channel which allow attachment of the landing gear legs.  The Mothership

developed a beef-up kit for the side skins (previously covered in a blog post) and a kit to substantially strengthen the landing gear attachment.  The kit consists of thicker aluminum and steel plates (shown in the pic at left) which go between the U-shaped bracket (U-1202) and the center section.  The T-shaped aluminum piece (T-1204V) on top of the center-section web is a drill guide with two #30 holes, which the builder was to use to match-drill through the web and the steel and aluminum plates below the web (U-1202C-B and U-1202D-B).  These #30 holes were then to be enlarged to 1/4 inch.  Happily, since I purchased my finish kit in 2016, the U-1202 parts already had the 1/4-inch holes drilled (my experience with match drilling holes has been less than enjoyable).  I only needed to drill the holes in the web.  The problem was that the holes I drilled in the web had to be perfectly aligned with the already-drilled holes in the other parts.  This wasn't an issue with the original beef-up kit since all the holes were drilled together and were automatically aligned.  Past experience using thin aluminum sheet with #30 holes as a drill guide showed that it's quite easy to "hog out" the original holes, causing the holes which are supposed to match the original to be off center.  I had visions of screwing up the center-section channel, which is the main structural member in the fuselage.  These 1/4-inch holes needed to be dead-nuts perfect.

To reduce the chance of misalignment, I ordered additional drill guides from Van's (they're cheap) so I could use stacks of three.  This allowed me to use a #30 transfer punch to accurately center punch the hole in the web and reduce the tendency for the #30 drill to walk off center during the drilling of the #30 pilot hole.  As instructed in the build manual, I used the reflection of the drill bit in the shiny drill guide to keep the #30 bit perpendicular to the web.  I then repeated the process with the 1/4-inch bit.

I won't know how well the holes align until I bolt on the gear.

Page 30-02: Fuselage pin stopper assembly

The fuselage pins are the heavy-duty steel cylinders which secure the wing spars to the fuselage in a fashion similar to that of glider wings, allowing easy and quick removal of the wings.  The fuselage pin stopper is the mechanism which keeps these pins from backing out in flight, an important task in

my estimation.  The stopper is a spring-loaded aluminum cylinder which must fit within the small steel tube welded to the main pin.  The spring is inserted, followed by the stopper, a hole within which has been tapped to allow the machine screw assembly to be threaded in, holding the whole thing together.  Problem was, the stopper was not even close to being able to fit into the said tube, the OD of the stopper being significantly larger than the ID of the tube.

I first tried fixing this by using  a pistol bore cleaning rod with a strip of ScotchBrite inserted in the slot rather than a cleaning pad.  The rod was turned in the ID of the tube with a drill.  This didn't remove anywhere near enough metal to make it fit.  I next created a poor-man's reamer by cutting the head off a 4-inch-long 5/16 bolt, then cutting a 2-inch-long lengthwise slot in the bolt

(shown in the third pic).  Into the slot I threaded a length of fine-grit 2-inch-wide sandpaper from my belt sander, wrapping enough around the bolt to make a tight fit in the ID of the tube (shown in the second pic).  I chucked the bolt in the drill.  The sandpaper looks too big in the picture because it unrolls when you're not holding it.  After lots of trial fitting and reaming it fits perfectly.  I cleaned the ID with the bore-cleaning rod (to the left in the picture) and swabs, then applied some Boelube and assembled the whole thing.

All of the other builder blogs I could find which discussed this spoke about the earlier system which required that a magnet be inserted in the stopper and it then be filled with epoxy.  The magnet was somehow used to tell the Skyview that the pins were secure.  That system apparently proved

problematic and was replaced.  Not sure how this new arrangement lets the Skyview know that everything is secure.  I suppose I'll find out in the near future.

The finished assembly is shown at left.  When the wings are installed, the machine screws allow the stoppers to be retracted and then snapped into a hole in a plate in the fuselage.  I doubt the pins would ever back out in flight even without the stoppers, but it would ruin my whole day if they did.

Service Bulletin SB 16-08-24 : Loose engine mount bracket rivets

Each one of the two upper engine mount brackets has 22 LP4-3 rivets attaching it to both the panel base and the upper firewall.  On two high-time RV-12s the aft-most six rivets on each bracket have been observed to be "smoking."  Two solid surfaces which are mechanically held in contact with each other and which are subjected to vibration can experience fretting corrosion.  Fretting corrosion

occurs when the two contacting surfaces experience relative motion, often of microscopic magnitude.  In the case of aluminum, the small-scale rubbing between mating surfaces produces fine aluminum power which quickly oxidizes, causing a smokey appearance around the rivet head.

Van's fix is to replace the offending LP4-3s with much stronger Cherry rivets.  This, of course, required drilling out the 12 LP4s in the engine mounts.

Some time back, Van's switched from the Garmin SL-40 com radio with separate intercom to the Garmin GTR-200 com radio with built in intercom.  Naturally, I had just installed the radio tray towers (the grey objects shown in the pictures) for the old radio when the announcement came out about the replacement, so those rivets had to come out as well (6 LP-4s total), so I did all the rivet removed at once.  The manufactured heads on the rivets were on the bottom of the of the panel base, so I originally started the process by awkwardly lying on my back in the fuselage drilling upward.  The drilling process in rivet removal is quite delicate, since enlargement of the #30 hole is to be avoided.  A clear view of the rivet head and a steady hand are required, both difficult to achieve in that position.  It occurred to me, somewhat belatedly, that this process would be much easier with the fuselage on its side, allowing me to sit upright to do the drilling.  Once again, I'm reminded of what a great decision it was to put off installing the tail cone as long as possible.  Without the tail cone, I have great access to the area between the aft bulkhead and the seats.  There's still lots of stuff to do there.

As much as I'd like to install the landing gear, I'm resisting that urge as well.  With the gear on, even without the tail cone, it'll seem much more like a real airplane.  I would no doubt be unable to resist the urge to sit in the cockpit and make airplane noises.  The downside, of course, is a lack of options for positioning the fuselage on its side while installing the many systems still to go in the fuselage.

Page 18-06 again: Flaperon Klöster Föken

Or, alternatively, Why I'm Re-ordering the Flaperon Skin

On several Van's designs including the RV-7 and RV-12 (unfortunately), a folded trailing edge skin is used for the control surfaces.  In general airfoil design, in order to facilitate the Kutta condition, the trailing edge should be sharp (not practical in the real world), squared off as is done with fast glass aircraft, or rounded, provided the radius of the trailing edge is less than 5% of the wing chord. I've

always considered rounded trailing edges to be an aerodynamic abomination no matter what the radius.  It offends me.  Van's did it right on the RV-9 which features a riveted trailing edge.  Coincidentally, the -9 is considered by many to be the sweetest-flying RV and it was the target for the handling on the -12.  Regardless of this complaint, having flown an RV-12 four times I can happily say that the -12 is the sweetest-flying airplane I've personally flown.  However, that doesn't mean it couldn't be better.

As discussed in the previous post, the build manual says to flatten the skins on the aft portion of the trailing edges using the hands, presumably thumbs, so that the skin is perfectly flat and parallel to the ribs.  I attempted this just prior to my annual week in Oshkosh (my 28th trip!) with limited success. Once I got to the show, I cornered an engineer from the Mothership (young guy whose name I can't remember) and asked him what I was doing wrong.  He explained to me that the way they do it at Van's is by using two blocks of wood, one on the upper and one on the lower surface, and squeezing them together with a large pair of Channel Locks.  This seemed to make sense, so upon returning to my NC home I attempted it.  The problem, and it turned

out to be a massive one, was that I misinterpreted the procedure he described to me.  I assumed that the wood blocks should be placed on the upper and lower surfaces just forward of the trailing edges, approximately where one might apply thumb pressure. Wrong.  The blocks should be placed on the trailing edge with the aft portion of the wood behind the trailing edge, hanging out in the air.  I found this out on another blog post, which re-posted the procedure from somewhere on Van's site, complete with pictures.  It turns out that this is the same procedure used to fix a "heavy" wing during flight test.  It was my mistake, but I wish Van's would publish a link to this procedure in the build manual, rather than just saying "flatten it with your hands."

Turns out I made several creases in the skin.  It wouldn't have hurt performance, and I tried to tell myself that I could build on, fly it without paint for a year or so as most people do, then build another left flaperon.  I was so unhappy with it that I drilled out several hundred rivets and removed the offending skin.  After (expletive redacted) up the left flaperon, I did the right flaperon with my thumbs (hadn't discovered the correct use of the blocks yet), and it turned out OK.  I may touch it up with the blocks later.  The new skins are on order.  They were surprisingly cheap.  

Update

It turns out that the aft flaperon skin can, indeed, be replaced while leaving the forward skin and aft ribs in place.  Upon completion of this, I discovered that the aft skins on each flaperon weren't lying flat as prescribed in the build manual.  I therefore constructed the tool shown to squeeze the trailing

edges, thereby flattening the skins.  I first practiced on one of the aft skins I had previously damaged, then proceeded to squeeze the trailing edges on both flaperons.  Success at last.

The "springs" connecting the two wood blocks are from a large binder clip.  The wood blocks have rounded edges with a large radius of curvature in order to prevent creasing the skin.

Pages 18-05 and 18-06: Flaperons

With the main part of both wings closed (hope I didn't leave any tools in there), the flaperons are the only thing left before starting the finishing kit (inventoried!).  The worst thing encountered so far is match drilling pre-punched holes in the aluminum flaperon leading-edge skin into the much harder stainless steel tube which is used as a counter weight for the flaperon.

My experience match drilling anything has not been positive, and this was by far the worst.  The #30 drill tended to "walk" in contact with the stainless, elongating the holes in the aluminum.  After this happened a couple of times I started drilling a #40 pilot hole first, then drilling the #30.  Worked much better.  I've always dreaded having to drill anything into stainless steel.













The next problem involved squeezing solid rivets such that either the manufactured head or the shop head would be seated against a curved surface.  I didn't want to machine a flat surface into the piece

since that obviously would weaken it.  The squeezer dies also wouldn't clear the flange.  The picture shows my options.  It seemed intuitively clear that the manufactured head wouldn't work at all on the curved surface.










I ended up grinding a flat surface on the cylindrical surface of the flat squeezer die (hard to see in the third pic) and putting a radius on the intersection of the two flat surfaces on the die (shown in the pic).  I also spaced the die out from the squeezer jaw with rivets.  Squeezing the shop head onto the

curvature worked fine.  The result is also shown in the third pic.  The shop head easily conformed to the curved surface.







The aft flaperon skins presented a new PITA right off the bat.  There were notes on the skin packaging warning that extreme caution should be exercised when unpacking in order to prevent damage, the reason being that these parts seem to be the most fragile I've encountered so far.  I suppose the flaperons need to be light, and therefore thin, to avoid flutter.  To make matters worse, the protective blue plastic was difficult to remove, especially on the inner surfaces.  And, for the first time in all the plastic removal I've done, a layer of sticky crap adhered to the aluminum when the

plastic was pulled off.  Somewhat belatedly, I discovered that if I apply the pulling force at right angles to the aluminum rather than at something between 135 degrees and 180 degrees as I was doing, the aforementioned sticky crap was not deposited on the aluminum.  Unfortunately, this was impossible on the inner surfaces.  Mineral spirits seem to get the deposit off with considerable effort.  What a waste of time.





Page 18-06 warns the builder to be certain that the tapered aft portion of the skins doesn't bulge out from the flaperon ribs but rather lies in a straight line against the ribs (see pic below).  Considering that this is matched-hole construction and the parts are riveted together, I don't see how the skins have any choice but to lie flat against the ribs.  Maybe I'm missing something here.  A revision in the build manual specifies a Cherry (stronger) rivet for the holes nearest the trailing edge in place of the usual LP4-3 rivets.  I suppose this is to positively

eliminate any bulging.  Unfortunately, I had to order the newly-specified rivets and am dead in the water until they arrive.

After pondering the bulging skin issue a bit more,  I think the manual is referring to the skin between the ribs even though it says "lie flat against the ribs."

Angle of attack probe -- thank you Joe Gores

Before closing up the skin on the left wing, it made sense to plan for and lay out everything needed to use the built-in feature of the Dynon Skyview which allows a display of angle of attack.  All that is needed is a pressure signal from a port on the leading edge positioned at a specific angle to horizontal.  All the hardware and positioning information is given in the thread shown below on Vansairforce.net.  This thread should be read in its entirety.

AOA link here

As suggested, I mail-ordered the parts from McMaster-Carr for the princely sum of about $15 for all the fittings, tubing, etc.  The hardware can't actually be installed prior to receiving the airworthiness certificate from the FAA, of course.  (Voice from the future: Van's came out with their own version of this with the hole in a slightly different place.  The slightly different placement is accommodated in the calibration.  Works the same.)

For the port itself, an LP3-6 rivet with the mandrel punched out works great.  A short piece of the same 1/8th inch ID tubing used for the static ports fits nicely over shop head, sealed and held in place by a glob of instrument-safe RTV. A 1/16th inch ID tube is coupled to the 1/8th ID tube and run to the wing root for later attachment to the Skyview. A mockup of these parts is shown in picture two.  The way it will be installed in the leading edge is shown in picture one.

The method that Joe Gores came up with and detailed in the link above can be used to locate the angular position of the port without actually having to measure any angles.  I'm sure that the calibration procedure can accommodate some angular position error in port location.  The early users of this method employed a common needle valve used for inflating basketballs (or deflating footballs if you're from New England) for the port.  By all reports, the hollow rivet works as well and is indistinguishable from the other few thousand rivets in the wing.

Some have suggested that the 1/16th ID tubing will fit into the existing grommets installed in the ribs for the wing-tip light wires (seen in the picture).  I couldn't make this work, even if I removed the string that we all ran according to plans to make is possible to run wires after completion.  There's a convenient route for the tubing against the spars, shown in the pic.  After closing the wing, everything can be reached with the inspection plate removed (the inspection hole can be seen at the lower right of pic one).

Both wings are now complete and are stored in the wing stand.  Work is proceeding on the flaperons.