Spars are complicated things—or at least they are when laminated out of various thicknesses and layers of aluminum attached to complicated spar caps. In previous installments of this series, we’ve covered the basics of engineering spars and what the various pieces do for wing strength. We’ve also gone through the detailed layout and basic assembly of a fairly complex aluminum spar for a Xenos motorglider. This month, we’ll look at the final steps—riveting the spar together.
The many changes in thickness along the “T” spar cap (covered in blue tape in this photograph) meant that the squeezer gap required frequent adjustment.
Most metal airplane builders are used to driving, squeezing, or pulling rivets that are 3/32 or 1/8 inch in diameter. These are fairly easy to set with simple hand tools, using manual force to pull blind rivets or to squeeze solid ones. Of course, many (including me) like to take the easy way out and use a pneumatic squeezer or puller, but they aren’t required. Shooting such rivets does require a pneumatic gun, of course, but in general, these rivets are easily managed.
Most spars, on the other hand, being the major structural component of the wing, have a significant number of 5/32- or even 3/16-inch rivets (that’s -5 and -6 for the experienced readers among you), generally solid ones, requiring considerable force to set. They can be shot, of course, or even pounded with a hammer and bar—but squeezing is always my preferred method of dealing with any rivet, if I can. You can get a more controlled set that way and more consistent shop heads.
The author (seated) and neighbors examine the specialized yoke and pneumatic squeezer before setting the first rivet.
The Big Squeeze
The average pneumatic squeezer can set -5 rivets quite easily, so long as it can reach them. Many built-up aluminum spars use rectangular bar stock as caps, and this makes for good, easy access to the rivets using a large longeron yoke. The Xenos spar that we have been using as an example here is a little different: The spar cap extrusions have a cross section sort of like a mushroom, and getting a good, straight shot at both the shop and factory heads can be difficult. Add to that the fact that the Xenos spar is long and takes two people to handle, and getting a good shot at each rivet becomes difficult.
The Sonex factory recommends using a hammer and backup bar to set these rivets, but again, there are a lot of them, and getting them consistent can be tough. So we took a more measured approach, calling on a friend who has the design experience and tooling necessary to build his own pneumatic squeezer yokes. Because of the force needed to set the -5 rivets, he fabricated a welded yoke consisting of multiple layers of solid steel to add the necessary stiffness so that it wouldn’t spread under maximum load. This made the yoke fairly heavy, and with the pneumatic actuator installed, it came in at close to 70 pounds—not exactly a handheld tool.
The answer was to set the squeezer on a base, and clamp that base to a small, but stable, workstand. The spar was supported on five roller stands (normally used to support wood going in and out of a table saw) obtained fairly inexpensively from a well-known inexpensive tool company. The spar could then move back and forth on the rollers, through the jaws of the yoke, and be lined up for each rivet in turn. Easy…right?
Well, conceptually, yes, and every rivet was set that way. But the fact that the spar web tapers from root to tip means that rivet lengths changed as we moved from one end to the other. Fortunately, the squeezer was designed with a screw-in set holder on the bottom, so it could be adjusted as we went. But it wasn’t as simple as that, for the spar incorporated clips for the inboard ribs, and every time we came up on one of those, we had a couple of extra layers of sheet metal to add to the mix, so the rivet length changed again, and that meant readjusting the lower set screw.
A make-your-own die extension was necessary to squeeze rivets where the spar cap interfered with the usual cylindrical squeezer ram. The die extension started as a machined cylinder of 4130 steel with a die “tail” on one end and a hole for the actual die tail drilled in the other. To give adequate clearance to the spar caps, the newly lathed extension was then put in the mill and a notch was cut in to give the appropriate clearance.
Making a Die Extension
As mentioned earlier, the outboard rows of rivets were centered on a line almost even with the edges of the flange, and that meant that lining up the center of the squeezer dies was very difficult because of the diameter of the squeezer rams and the dies themselves. This was solved with a die extension, a machined cylinder of 4130 steel, with a die “tail” on one end and a hole for the actual die tail drilled in the other. These were done by chucking a slightly oversized 4130 rod into the lathe and cutting it down to final size. But this wasn’t enough to give adequate clearance to the spar caps, so the newly lathed extension was then put in the mill, and a notch was cut in to give the appropriate clearance.
This worked well, up to a point. Squeezing the large rivets took considerable force, and every time this force was applied to the non-symmetrical die and extension, it wanted to bend the extension a little bit. This meant that the extension was only good for a finite number of rivets before there was enough misalignment to make the rivets ugly. So we didn’t just make one of the notched extensions; we made quite a few before both spars were done. In fact, I got pretty good at using my lathe and mill before we were finished! But with a steady supply of material and some quality time making chips, getting to the end was never in doubt.
Is That Really the Right Rivet?
Because the spars of the Xenos overlap under the seats, the aft face of the root of one spar and the forward face of the root of the other had to sit flush against each other. Not only did the rivet length have to change, but the type of rivet changed as well. Whereas round-headed universal rivets were used for most of the spar, flush rivets were necessary on the overlap area. This meant that all of the countersinking and dimpling had to be done in the setup stage, before any riveting had begun. Careful review by multiple eyes was necessary before the first rivet was set, just to make sure that no holes and dimples were missed.
Liberal annotation with markers helped ensure that the correct rivets were used in each hole as the length and type of rivet varied throughout the spar.
Because of the uniqueness of so many rivets, we also found that it was necessary to go down the length of the spar with a Sharpie and a copy of the drawings and mark each area with the length (and type) of rivet to be used for each location. There were some holes near the root that were for later alignment purposes and didn’t require rivets; these were taped over (on both sides) to prevent an accidental rivet placement. The design also includes a number of AN3 bolts near the root to hold the thickest portion of the web together; these each had a bolt, nut, and washer callout on the drawings and had to be done separately from the riveting. And with tiedown points and bellcrank mounts along the spar as well, close attention and check marks on the drawing were required. Getting into “production mode” and doing the same thing over and over was a great way to make a mistake.
One last complication is that rivets should generally be driven with the manufactured head on the side of the thinnest material to avoid the puckering that can happen when you form a shop head on thin material. The way the Xenos spar is laid out, the thick/thin layers change sides (several times) from root to tip, so not only do you have to swing the spar end for end to reach both sides with the squeezer, but it has to be flipped as well.
A second person nearby was nearly essential while squeezing the rivets on the Xenos spar to help lift, spin, and flip the spar as needed. The second person also helped ensure that the proper rivet was used for each hole.
Does all this sound complicated? Well yes, it is. It requires great attention to detail, a lot of patience, and quite a lot of time to get it all right. We aren’t an airplane factory with eight-hour days, so it probably took close to two months of intermittent sessions, working a couple hours at a time, to get it all done. Because of the size, it is often a two-person effort to lift, spin, or flip the spar. But because of the intricate setup for the different length rivets, one of the people is spending most of their time waiting for the other to set up and squeeze things. So it is a good idea to have other tasks—like deburring wing ribs—going on at the same time.
Quality control of wing spars is important, so when you think you are done, it is important to go over it all one more time, with drawings in hand, and check every rivet to make sure it is set properly and in the correct direction. Make sure no holes have been missed. And check to make sure that nothing has been scratched, bent, or gouged, as well. This is a good time to enlist another experienced builder to check your work; they can dispassionately evaluate the work and recommend things that need to be redone. Clean off all of your notes and marks before the inspection, so problem rivets can be circled and marked. Check your ego at the shop door before this process begins; it’s about getting it right, not you being right.
Only after the quality checks—and any necessary rework—are complete should you begin to tear down your spar fabrication workshop and supports. You will have spent a great deal of time getting to this point, so take the extra time necessary to get it right.
With both the right and left wing spars completed for the Xenos, the pieces now lie in their final dihedral position and ready for final mating.
Is It Worth It?
Spar construction is rewarding in the end because you will better understand the overall design of your aircraft by participating in the construction of the heart of your air machine’s wings. It is not going to be cost effective, however, unless you set the value of your time at next to nothing, so it is reasonable for builders to pay the extra money for pre-assembled spars if what they want is a flyable airplane in a shorter period of time. It is hard to beat the quality and workmanship of someone who has done a lot of these, and yes, you’ll make mistakes—some of them potentially expensive—by doing it yourself. In the old days, every builder did their own spars, either out of metal, wood, or composites. This can connect you to homebuilding’s roots, if that’s your kind of thing.
We hope that you have enjoyed and learned from this series on spars. The heart of your wing is the essential component that keeps you safe, whether in a glider or flying aerobatics. Understanding how they are designed and built is a great piece of the overall puzzle that makes up the building of an Experimental aircraft. And remember, this is all about recreation and education. Building a spar, for someone who likes to do fabrication, will provide plenty of both.