02 March 2014

Empennage: Horizontal stabilizer. Factory error on rear spar?

First, let me extoll the dedication of Van's.  They went beyond what I would have expected in looking in to this issue for me, even though their diagnosis wasn't correct (let's be fair, they can't actually get their eyes and hands on my parts).  They are going to send me a replacement rear spar and doubler.  They are a great company to work with and are customer focused!

Previously, I had indicated that the rear spar on the horizontal stabilizer was grossly misaligned with the stab assembly.  In fact, it was simply not possible to rivet it in.  It seems it wasn't related to the bow in the spar.  Other -14 builders I consulted with indicated that their rear spars also had a prominent bow and it wasn't an issue in the final assembly for them.

An example of what I was up against is shown below.  Notice how the trajectories of the unset rivets are off (in fact the lower left-most one can't even fit through the hole) and the skin pushing away from the spar maximally at the red line.  The associated dimples won't sit in their countersinks either.  What's interesting was that the misalignment was not progressive, but rather discontinuous at the red line and happened at the same four (symmetric) places along the rear spar:  Between the 26th and 27th rivets when counting from outboard on all four corners.  

After spending several hours trying to determine the cause of the problem, including a visit from a tech counselor, a RV-9A builder, a call to Van's and several emails with the latter, everyone concluded the problem was most likely a twist induced in the assembly when it was riveted together.  However, I wasn't convinced that the stab was twisted.  Since the issue presented itself as a discontinuity in rivet alignment, I couldn't wrap my mind around the problem being my assembly having a twist.  After all, wouldn't a twist manifest with a progressive misalignment?  So, I was quite certain that the spar was mis-drilled at the factory, but I approached the problem with an open mind since I'm often (well, frequently...actually, just about all the time) wrong.

So, to chase down the alleged twist, I tried 
  1. Cleco'ing in various orders to see if the misalignment could be "chased" in one direction or another.  It couldn't.
  2. Applying forces to the assembly at various points to slip rivets in with a straight trajectory.  That wasn't successful (and having rivets in sheer is not good)
  3. Pulling a string line through the rib tooling holes in the assembly to check for twist.  It looked pretty good
  4. Van's suggested seeing if anything changes when the spar is inserted backwards (top-to-bottom).  Problem persisted
  5. The great folks at Van's even tried a few mockups with parts from different batches and couldn't reproduce my problem.  They continued to feel it was a twist and suggested removing some rivets and reworking the assembly
At each point in the above diagnosis progression, I found no evidence of a twisted assembly.  Ultimately, I requested the rivet spacing for the parts.  The standard spacing is 1.125".  Sure enough, right where I identified the misalignments on my spar (each of the four places I marked), I found one pair of misdrilled holes at 1.175", 0.05" too big.  The corresponding holes in the skin did not have this additional 0.05" spacing (see this later post which expands on that statement).  This explains the issue I had and why it manifested as a discontinuous affair rather than a progressive one.  Images below illustrate my findings.

In retrospect, I should have just measured the holes as soon I suspected that was the problem.  However, I had it in my mind that the holes had different spacings, like the J-stiffeners on the wings and that thought locked me out of just measuring the holes.  Had I just made the measurements, I would have saved a lot of time.

If you're curious why the spar's countersinks appear not to have been primed like the rest of the part, it's because I realized that my countersinks were slightly shallow, so I shaved off a few thousandths to bring them to the right depth.  I had planned to spot prime them if the spar could align to the skins.

Visit this link for the next entry, 8-May-14, on this issue.  It includes the reason why the spar was mis-punched at the factory.

Wings: Fuel tanks redo. Riveting sender nutplates.

Because I'm redoing my fuel tanks, I'm really only posting progress reports rather than additional details.  See here for why I'm redoing my tanks.  Technique details on the original fuel tanks are found in their associated posts.
There are five nutplates for each fuel level sender that must be attached to the inboard rib of each tank.

Six of the ten rivets holding these nutplates are easily reached using a squeezer with a standard 3" deep yoke.  The forward most 4 rivets must be reached with a deeper yoke, such as a 4".  Barring that last yoke, these four rivets can be back-riveted with a dumbell bucking bar.  Just make sure the curvature of the dumbell doesn't overlap that of the rib, otherwise the rib's "bump-out" will be deformed.

Setting up the bucking bar on the manufactured heads.  The cleco pliers is to balance the rib so it doesn't fall over.

Here the bucking bar is set so the rivet in the upper right is ready for bucking.


In the image above, you can see my capacitive plates in the process of being formed.  And on the rib, you can see the additional holes for the capactive plate (fabrication & installation on original tanks) BNC connector as well as for my low level optical fuel sensors (spec'ing & installation on original tanks).

I actually didn't remember how I set those forward four rivets on my first tanks.  I had to review my fast-motion video from July of last year to see how I managed those.  See here for information on that setup.  Though I don't post the videos, I do post images from that camera from time-to-time (they're the ones with the bird's-eye view perspective like the above).