Disclaimer

This blog is for entertainment purposes only, and is not meant to teach you how to build anything. The author is not responsible for any accident, injury, or loss that occurs as a result of reading this blog. Read this blog at your own risk.

Monday, January 30, 2017

Ch 10 - Canard - Part 9

Locating the inner elevator hinges (7.0 hrs)

It might seem a bit early to start worrying about elevator hinge locations since I haven't skinned the top of the canard yet, however I need to bury a number of hard foam inserts into the canard where the hinges will go, so I have to make sure I know fairly well where that will be.


Normally, the hinge locations would come straight from the plans, but my longer and wider than specs nose created a few additional challenges, especially when you factor in my choice of Cozy Girls offset torque tubes. 


GU canard hinge location

Basically, I will be using the original plans for generic guidance, the Roncz plans for more in-depth directions, and the Cozy Girls instructions for more specific assistance. These last ones often refer to the Cozy manual which I don’t have.


Roncz canard hinge location


Saying that elevator installation geometry gives me headaches would be a gross understatement, and at this stage I am not even sure I understand it all myself. Thankfully my buddy Wade has already assembled his elevators to the canard, and made a short video of how it is all supposed to go together. I will link here with his permission.





Wade's elevator mounting video




Thanks again Wade.

In the video you might have noticed that Wade used only six hinges to mount his elevators, four of them are externals, while the two inner ones end up inside the fuselage. His outer hinges match the Roncz plans’ locations, while the inner ones were moved due to his slightly larger nose. Mine will have to be moved as well. Unfortunately for me his fuselage is slightly narrower than mine, so I won't be able to replicate his hinge measurements, and I will have of figure out their locations myself.

Because the original torque tubes rotate about an offset axis, they need a small slot (a bigger hole really) cut in the fuselage to be able to move. This historically leaked cold air into the fuselage during the winter. The Cozy Girls offset torque tubes eliminate the need for the slot, which allows for a tighter seal with the nose structure, minimizing cold air infiltration in an area of the plane that is not usually heated.

I will concede that the change in hinge location due to the non-standard nose is further complicated by my choice of offset torque tubes, but the reward of warmer feet in the winter appealed to me.



The original torque tube (in back) would have required a slot in the fuselage to be cut for clearance.

Checking how these offsets fit on the fuselage of my Long EZ (remember these were designed for a Cozy), it was clear that they could be mounted in a number of ways.


Offset torque tube centered on fuselage. Note the the wide gap.


Offset torque tube biased inward to bring the elevator closer to the fuselage


Talking it over with Wade, I decided to bring the offsets as much inboard as possible, in order to bring the elevators closer to the fuselage. I figured I could leave about 0.100” (2.5 mm) of play to account for finishing the skin, and then still have a little gap. 

All I needed to do at this point was to measure the distance between the slots built into the offsets, as this would represent twice the BL of the hinges. The problem was that handling two torque tubes, two thickness gauges, and a measuring tape long enough to take multiple measurements required at least six hands.

At least four hands short as usual, I decided to use a PVC pipe scrap leftover from the anodizing bubbler build, and two yet-to-be 3D printed adapters to hold the offsets in place. I could then test at will, shortening the pipe as needed, until satisfied with the results, then I would bring everything to the bench where more precise measurements could be taken.

The idea had a great potential to improve accuracy, and simplify my measuring task, however a little CAD work was necessary.

I started by redrawing the offset torque tubes, then modeling inward from there.



Drawing up the part gave me a much better understanding of this complex piece. The Cozy Girls rock!


The left side of this adapter fits the torque tube, the right side fits the PVC pipe.

Adapters and torque tubes

Adapters and PVC pipe

This is what I was working to achieve

With the adapter files transferred to an SD card, I fired up the 3D printer, and let it rip.



The easiest way to print this is standing on edge

This way of printing leaves smooth sides already the proper size

Snug fit
 
Once this assembly is properly fitted to the fuselage, this will be the distance I need to measure.


Unfortunately, the low-density infill adapter proved too weak to carry any weight, and failed before I got the chance to position the assembly on the fuselage and take measurements.



Anatomy of failure

So, I maximized the infill on the second print, which unfortunately lasted only moments longer.



Adhesion between contiguous ABS layers proved insufficient

The shear force acting on a plane parallel to the layer deposition plane, made the adapters unsuitable for their intended purpose. 

While I could have printed the adapter horizontally in the 3D printer, this would have made it much rougher, and thus harder to fit on the tubes. Instead, I decided to redesign it slightly, and insert a 0.25" (6.3 mm) steel pin into them after printing.



A different kind of "composite"

Because this would be an interference fit, I added a small hole through the bottom for air pressure to escape as the pins were forced in.



Broken dreams and "empty" promises

I ended up reaming the center hole 0.001" (0.025 mm) undersize

A few gentle taps with a plastic mallet persuaded the pin

Steel pin bottomed out

The changes to the adapter worked out well. Although this process took longer to bear fruit than I had expectedprecisely measuring the hinge distances became a total non issue.



Finally back in business

After a few PVC pipe shortening sessions on the belt sander, I reached a happy distance.

Using gauge blocks (on both sides) to quantify elevator clearance

The distance between the slots turned out to be 20.0" (50.8 cm) at the outer edges, and 19.6" (49.8 cm) at the inner edges, so I chose 19.8" (50.3 cm) as slot centerline to slot centerline distance, which gave me a hinge BL of 9.9" (25.1 cm).

So much work for such a tiny number.




Sunday, January 29, 2017

Ch 10 - Canard - Part 8

Trailing edge peel-ply removal (1.7 hrs)

Before securing the canard to the table once again (right side up this time), I decided to take care of a few operations on the trailing edge, while I could still slide it around the table for better reach.

The main objectives were to remove the foam “fishtail”, and the embedded trailing edge peel-ply right below it, then sand the trailing edge to a smooth transition.

I used a small saw to cutoff the bulk of the fishtail.


Removing the fishtail


Next, I dug the foam out until I got to the pure epoxy crusty layer covering the embedded peel-ply/staples layer.


One of the staples ripped out of the peel-ply

Clearing this much epoxy crust took nearly one hour. This was some slow going!

I’m sure someone else must have already figured this out 30 years ago, but not knowing about it, it had to be rediscovered all over again. So, acting on my friend Ary’s suggestion, I didn’t apply any micro-slurry over the peel-ply (before fiberglassing)


Back in time to show you that no micro was applied above or below the peel-ply before fiberglassing

As a result of this decision, and with the canard sitting right side up now, there was nothing but pure epoxy between the peel-ply and the fiberglassed bottom below it. Ary’s theory was that this should make the peel-ply removal much easier than if it had had micro-slurry.

Because some of my friends took up to three days to clean up this area, I was willing to try anything that had even the remotest chance of making progress easier.

Skipping the micro turned out to be a great suggestion after all, and ripping peel-ply and foam became nearly a pleasure.


While the first inch took nearly one hour of chipping hard epoxy, the next 4" took only seconds.

Perhaps I should show the video I made to give you a better sense of the lack of effort involved using Ary’s idea.





Insert peel-ply video




With the peel-ply out of the picture, I marked where the fishtail had merged into the airfoil with a marker, and gently sanded the foam aft of the black line smoothly into the trailing edge.


Long black mark is where the fishtail began

Sanding until the black line starts to disappear

One additional benefit of not using micro-slurry was that no additional sanding of the fiberglass trailing edge was necessary, due to the textured surface left by the peel-ply, and the absence of micro.



Tuesday, January 10, 2017

Ch 10 - Canard - Part 7

Flipping the canard (10.1 hrs)

With the bottom of the canard completed, the plans instruct to separate it from the jig, and flip it over in order to start glassing its top surface. Before that could be done though, a new jig would have to be built on top of the canard’s bottom surface (currently facing up) to avoid disturbing its carefully aligned span. 

The new jig should consist of a 130” (3.3 m) long straight pipe1.25” (3.2 cm) diameter fitted to the trailing edge, and four wooden steps bonded to the canard’s bottom.

The straightest longest metal pipe I could find was a few inches short of ideal, but lacking better prospects I went with it anyway, and attached it with Bondo to the trailing edge of the canard.


Plain automotive Bondo

Small dabs of Bondo applied to the trailing edge

Metal pipe attached to the trailing edge

The plans dimensions of the four wooden feet were 1”x4”x8”.


Plans' instructions on how to set up the new jig

I didn’t have a one inch thick board from which to cut them, but I had a leftover 2x4 beam laying around, so I used that instead, and changed the design slightly as well. 


I carved a small foot out of the block to make it sit better on the canard

I took great care to make them as close to identical to each other as possible.


Using fences and stops to help with repeatability

Next, I made a “movable jig” to help me line up all four pieces at the same 2” distance from the leading edge.


The two inches aren't critical, but setting all feet at the same distance will help later.

With this homemade fence, some shims, a level, and a speed square, I positioned the wooden pieces on the bottom of the canard, then used hot glue to set them semi-permanently.


Setting the leading edge distance, and shimming to level.

Making sure the foot is not crooked

To visually check that all the wooden feet ended up at the same level, I once again pulled out the laser and aimed it at their top edges. I figured I could sand any differences down to the laser line, but I lucked out, and all edges lined up on the same plane.


Laser hitting the very top edge of the wooden feet

Same thing on the opposite end the canard

After a big sigh of relief, I pulled out my 8’ level to check the feet once again manually.


Spanning the first set of three wooden feet I detected no rocking, and no gaps with the middle foot.

Middle foot touching the beam, while measuring with two levels.

The second set of wooden feet felt just as solid as the first

Middle foot also touching, and zero leaning.


At this point I had done all I could, and the only thing left was to rip the canard off the former jig and entrust its straightness to the new jig. So I pulled up on the leading edge, which came off the jig without much trouble, and flipped the whole canard over onto the unused part of table, letting it down gently.


Canard is finally off the jig

To my great surprise and relief, the whole ensemble felt solid as a rock, and damage to the foam was minimal. Very satisfied with these results, I put the canard back on the old jig to do a little sanding on bottom surface of the leading edge (I should have thought of this earlier) in preparation for further glassing.


Peel-ply did most of the work for me, but there were a few strands left and a little rough edge in a few spots.

The final step was to sand the leading edge fiberglass to a feather edge.



Feathering the fiberglass edge avoids creating a bump along the leading edge

To accomplish this more precisely, and comfortably, I discarded the old jig, then I stood the canard vertically on the back end of the new jig, and screwed it to the table.


I used foam scraps to support the trailing edge

Four screws secured the whole thing to the table

Using the sanding block I removed the least amount of glass needed to blend the bottom of the canard smoothly into the blue top foam.


In general, when you are comfortable you do a much better job.

Leading edge smoothed

I tried very hard not to touch any of the blue foam with the sanding block

Later I removed the screws, and sat the canard right side up on the table where it would receive a little more preparation before further glassing.