Step 10.7, Roll bar

I don't know if I was having as much fun as my puppy, but it sure was good to get back to some basic construction.  In this post I am fabricating the roll bar.

The roll bar is designed to protect the noggin in the unlikely event of an unplanned inverted dirt landing.  The basic construction is a curved U-channel of pressed aluminum spaced apart by narrow strips of 0.063" sheet aluminum.  Here are the basic parts:

Some access holes are cut into the aft sides at the base and then the parts are laid out on the bench where the splice plate can be match drilled and then riveted to the center of the arch.

A splice plate is fabricated for the aft piece as well.  The aft piece also gets a bit of angle to mount the roll bar support that ties the top of the bar to the top to the fuselage.  This is probably thought to prevent the bar from laying over -- With any luck, I'll never have to find out if it is up to the challenge.

The narrow strips are cut to length and and then match drilled to the forward arch piece

Match drilling.

And riveted.  The inside strip is next.

Followed by the other half of the roll bar to complete the tube.

There are a number of flanges that are fabricated from various thicknesses of angle that are needed to attach the roll bar to the fuselage.  Carol has come down to bang out a few..

Here are two that will hold the outside of the base.

The bar is canted aft so the attachment flanges are cut back to reflect that angle.

The bottom of the bar has to nest within the inside of the angle brackets, so the corners are radiused with a Scotch Brite wheel.

Here the two brackets on each side are shown clamped in place prior to match drilling.

Oops.  The L-shaped brackets are the wrong way around.  The smaller of the two should be on the bottom.

That's better.  I should have known something was wrong when I felt a blissful surge of intuitive correctness about the assembly.  That just never happens.  Oh well, better off correct than aesthetically pleasing.

The U shaped channel at left attaches to the top of the roll bar and keeps it from folding over.  I have elected to add a bottom piece to it to give me a place to mount a baggage compartment light and for a more finished appearance.  I found some left over 0.025 sheet aluminum that was large enough and the thickness seems about right.  I'm going to mount the cover inside the U channel so that I can put the nut plate fasteners on on cover instead of inside the channel.  This will probably be easier although either way is do-able.

Now I'm back to the L brackets.  The outside piece has to be tapered down to match the curve of the fuselage.  I start out with the Vixen file because there is quite a lot of material to remove.  It turns out that on this particular piece, the filing is making a very loud high pitched noise.  I decide I'd better move it over to the belt sander before my puppy files a grievance.

I finally get the taper right and begin the final installation with counter sinks for two attachment screws.  You can see the taper in the edge of the counter sink.

Then the U channel is attached to the roll bar.

Then a quick run through the pait booth.  Here is the roll bar drying on a pile of interior panels.  

Here is the custom bottom piece.  I put it in place temporarily so that I could get the proper spacing on the two rivets on the sides.  The top is also riveted to the top skin.

And now...  The roll bar is finished.  Painted and bolted down.  Ain't it purdy.

Step 10.6, Test fitting H and V stabilizer

The cold weather has once again returned to Rough and Ready.  It was beginning to look like we were about to skip winter.  Now it's beginning to look a lot like Christmas --  That is, what Christmas would look like were I to live anywhere but California.

The following photo pinpoints for the first time, the exact cause of Global Warming.

Ahem.. Moving on now, before the CO2 police show up and arrest me for heating the Universe, I begin by retrieving the completed horizontal stabilizer from the loft.  This is the first time I've looked at it in a year and a half.  It's not too bad, but I can see that my riveting skills have improved some.  No show stoppers though, so I proceed with the attaching the HS.

Bolt holes must be drilled through the longerons and some support angle in the tail cone and through some spacers that sit on top of the tail cone deck plate.  I found this to be a little tricky because I was convinced that the best way to drill this was from the bottom with an angle drill.  I don't remember why I got stuck with this particularly stupid idea, but that is how it went with the first two through the longerons.  With your arms through the access holes holding and positioning the drill you can't see where you're trying to drill.

I finally got a little smarter and finished the holes through the HS forward spar located with a strap duplicator. Going this route, I could be sure the the holes would be plumb because I could see the drill.  Before the holes are drilled the HS is measured against fixed points along the fuselage.  When the tips of the HS measure the same on both sides the the horizontal stabilizer is straight and one should drill at once before anything moves.

Here Carol is helping me to measure.  She enforces the measure twice before drilling rule.  I get that mixed up all the time.  More on this later.

And then we move on the the bolts at the HS rear spar.

And here it is all bolted down.

So now I move on to the vertical stabilizer.

The vertical stabilizer connects to the fuselage at the stabilizer's rear spar and the front of the stabilizer to the  horizontal stabilizer's main spar through the F-781 attach plate.  The F-781 comes from Van's  as a flat piece of 0.063" aluminum stock.  It is already cut to size, but the edges need smoothing and some holes have to be match drilled with the horizontal stabilizer.  There is also a bend at about the midpoint to match the rake of the vertical stabilizer.

The F-781 attach plate has been the subject of much discussion on the internet forums because of some cracking of this part discovered in a few aircraft.  In addition, one or two crashes have been linked to the loss of the vertical stabilizer.  There is at least one company producing an aftermarket flange to beef up this area.

With all of this in mind, and with meticulous attention to detail and the fine craftsmanship of an elderly, but still quite sharp Swiss watchmaker, I set out to produce the best F-781 part yet committed to aluminum.  It would the object of admiration across the aviation industry and would probably make the cover of "Aviation Week and Space Technology".  I will be astonished if a finished piece this magnificent fails to catch the attention of NASA or Jay Leno or something similar.

Naturally, with plans this grandiose, the trouble will begin early.  After the F-781 is match drilled to the HS, I fear that I may have clamped it in place incorrectly prior to drilling.  Then, I decide that the holes actually are offset too high, but I still hold out hope that when the holes are drilled to final size the offset error will magically disappear.  So I re-position the F-781 and drill to final size.  In the photo below, it's an obvious FUBAR.  

This part is definitely scrap, so I trace the outline onto another piece of .063" to make a replacement. The hole locations are transferred and I'm back in business.

And here are the two parts together.

To complete the new part I have to put the bend in place.  It's a simple matter to match the old one if both are placed in the brake.

After this experience, I'm considering changing my long held belief that one should drill twice and measure once.  So here is the completed F-781 in place.  If the longitudinal dimension of the vertical stabilizer's forward spar requires it for a good fit, Vans allows a spacer to take up the slack.  I was able to reuse the miss drilled F-781 for this less critical purpose so, all's well that ends well.  (This may be the only builder's blog to quote Shakespeare).

I move on to attaching the read VS spar, which requires more match drilling.  Here is Carol holding the vertical stabilizer prior to drilling the rear spar.

And now the four bolts at the bottom of the rear VS spar.

Transferring the hole locations from the tail to the rear spar was a bit of a head scratcher because of the very limited space inside of the tail. It made it impossible to match drill from inside.  Transfer by paper template solved the dilemma.

And here are the four bolts installed.

And so here we are with the vertical and horizontal stabilizers attached for the first time.  I can't attach the rudder as long as the plane is on the rotisserie, so that will have to come later.  At this point there is nothing left to do on the HS and VS but take them back off in preparation for painting.

Step 10.5, Elevator push tube

There are a lot of odds and ends that need doing as I near the completion of the fuselage.  I feel some pressure to get these things done quickly so that I will be ready when the 4th and final kit arrives.  First, I need to thread the rudder cables through the snap rings in the bulkheads from the tail all the way up to the rudder pedals.

Right away I find that the cable ends will not fit through the snap rings.  This is just the kind of thing that annoys a builder with time pressures.  Well fortunately, the internet comes to the rescue once again.  I discover that the proper response to this calamity is to squeeze the snap ring into an oval shape big enough to pass the cable end.  Oh, Ok.

Then I go on to the elevator push tubes.  The push tubes connect the control column to the elevator.  There are two tubes in a linkage that runs under the seats then under the baggage compartment and then on to the tail.  The elevator bell crank which is just behind the baggage compartment, connects the two tubes in the middle.  The bell crank reverses the linear direction to the first tube so that when the second tube is connected to the elevator it's pushing and pulling in the right direction.  The second function of the bell crank is to change the geometry of the tube such that it is offset from the tail cone floor which is tapering upward toward the tail.

The push tubes are just aluminum tubes with end fittings for connecting the ball-end bearings.  The first step is to drill six evenly spaced holes around the circumference.  I used masking tape to measure the hole spacing and then wrap the tape around before drilling.

Before riveting the ends on, the inside of the tubes are coated with primer and allowed to dry.

 Now the big push tube.  Same procedure as the first, but this one has eight holes on each end.

A trip the the paint booth and the big push tube is ready.

Here is the bell crank with the small tube that runs along the floor already attached.

And here is the second tube connected to the top of the bell crank.

And there is goes all the way to the tail.

The big push tube connected to the elevators.

Finally, I bought some fairings for the rudder cable where it exits at the tail.  They come almost ready to go.  Just clean up the edges and measure out the rivet spacing.  The bottom row is too close to the J-stringer to get a bucking bar in place.  Rather than fight with it I'll just use flush pull rivets on that row.  That's the rudder control cable exiting the fairing to the right.

Next time I'll be test fitting the horizontal and vertical stabilizers.

Step 10.4, Test fitting the wings

The wings attach by bolting the wing's main spar into a slot in the fuselage's main spar carry through structure.  The main spar and the spar carry through were pre-drilled at the Van's factory and are delivered as a matched set to insure perfect alignment for the close tolerance bolts that will join the two pieces.

There a few important parameters regarding the wing attachment which are set when attaching the wings. First is the dihedral angle.  This is the amount that the wing tips are higher than the wing root.  Dihedral helps keep the top side of the airplane pointed up.  On the RV7, the dihedral angle is set by the predrilled holes in the spar attachment so nothing needs to be done on that.

The next parameter is the sweep angle which is a measure of how far back the wing tips trail the wing root in the horizontal plane. On the RV7 the sweep is 0 degrees, so the leading edge of the wing forms a 90 degree angle with the centerline of the fuselage.  The sweep is fixed by the rear spar attachment which is not yet drilled.

And the last parameter is the wing incidence.  Incidence is a measure of the angle of the wing (chord line) relative to the fuselage.  Incidence of the wing is set by the designer such that at cruise speed the fuselage will be in a level attitude.  Like the sweep adjustment, the incidence angle is also set by the location of the rear spar attachment.

The basic idea is that the sweep and incidence are both adjusted and then the rear spar attachment is drilled and bolted to lock in those parameters.

This will be a test fitting of the wings because they will have to come off again for paint.  Before moving on to the wings I took time out to install the control column which had been previously assembled.  Because I had made the tops of the four inside ribs removable, the control column could drop straight in with out any struggle whatsoever.  The photo below shows the ribs with the top sections removed.

I did have to disassemble the column pivot and move the bolt head to the opposite side for clearance.

After the control columns were in place, I painted some of the interior that would not be covered with fabric.

Then I began the process of fitting the wings.  First, checking the level across the longerons and then along the longerons.

 At this point I realized the it wasn't necessary or desirable to level the plane before fitting the wings.  Oh well, at least now I'm familiar with the procedure.

Guess what?  I have no pictures of the wings going on because everyone involved had their hands full.  On the right wing I had the help of Carol and her nephew, Cody.  A few days later, Rob and Tony from work came over to lend a hand on the left wing.  In both cases, the fit was very tight, but an application of Boelube to the mating faces helped some.  I used hardware store 7/16" bolts with the threaded end ground down to complete the temporary connection.

Once the right wing was one done I set out once again to level the airplane.  The reason we want to level the airplane is that the wings must be attached perpendicular to the fuselage.  To accurately measure this angle a right triangle formed by the leading edge of the wing and the centerline of the fuselage is transferred to the floor, which as it happens, is a mostly flat surface good for measuring.

The first thing is to find the centerline of the fuselage and then transfer that line down to the floor.

The self leveling laser comes in handy here.  I didn't completely trust the laser though, so I checked its accuracy with a plumb bob.  It was spot on according to gravity, which I reluctantly accept as an authoritative source.

The next step was to measure string a line across the floor directly beneath the leading edge of the wing. Once again the plumb bobs find the line.  This line will not intersect the center line at exactly 90 degrees if there is any sweep in the wing (not perpendicular).

Then I measure where the extended leading edge line crosses the centerline and take a measurement of that line and of the length of the centerline from the leading edge back to the tail.  Now having two sides of the right triangle, I find the length of the hypotenuse via Pathagoras and draw a line on the floor at that length.

Now here is the tricky part.  I have to move the wing so that tail to wing tip matches the length of the hypotenuse just calculated and then reset the leading edge line.  With the line reset, I remeasure the lengths of the leading edge and centerline and find that they remain unchanged -- at least to the accuracy of my ability to measure. Considering the lengths of the sides and the precision of the tape measure, the calculated angle should be within a few 1/100s of one degree.  Ok, Close enough, moving on.

Now the sweep is set, but the incidence must also be set at the same time.  

Van's has this high tech method for setting the incidence.  A premeasured block of wood setting on the rear spar forms the incidence angle relative to the main spar.  Remember that the fuselage is leveled before starting this process, so the procedure is actually true measurement of incidence.  I don't have a lot of confidence in the precision of a bubble level, but after turning the level end for end a few times I was convinced that the level was good enough.  The actual incidence is a not nearly as important as getting both wings set the same.  Since I used the same block and the same level and most importantly, the same two eyes, I feel that the two wings will be pretty close to the same angle.

Now that the sweep and incidence are correct, a series of increasingly larger holes are drilled through the rear spar attachment ending with this 5/16" reamer.

And there it is.  Let's hope that both wings get set the same.

Now do the other side.

With a little help from a jack.

Finally, check the fit of the fuel lines and add the vent lines.  

And here is is.  Wings on for the first time.