W14 flush riveting

[scotttro]

I need some clarity.

I’m preparing to flush-rivet the main spars, so I’m on W14. As I know nothing about riveting, I have been familiarizing myself with the whole countersink/dimple/flush rivet process. Just figuring out how to consistently *measure* countersinks has been a bit of an odyssey itself.

W14 says to countersink into W10-03 to 19/64”, or 0.297”, a couple of builder websites agree, and my CAD-type analysis showed 0.298”, so you could say “a little under .3” is a real solid number. But when I drive practice rivets at this diameter (and I’ve driven quite a few at this point), they stand proud 5-9 mil or more. As I increase the diameter, they stay proud until I get to .320 or so, where I get nice flush rivets.

I’m back-riveting using a 3x gun (as I have no access to flat concrete), and the dimples were made using both the Sonex simple dimple tool as well as a modified vice-grip with Avery dimple dies. The bucking bar is a 25 pound chunk of machined-flat steel.

I seem to have a flawed technique going on here, but I’ll be darned if I can figure what. The Google and the YouTube aren’t of any help. Perhaps some step I’m missing? This is a stage I really don't want to botch.

Scott

[lstinthot]

Here is a reply to a thread I started with the same problem. Turns out good tools make good dimples. We bought the countersink bits and used a micro stop to get the proper depth. Ended up having to buy more rivets due to so many practice rivets used. I also recommend starting at the wing tip end to practice on regular head rivets, by the time you get down to the countersunk ones you will be more comfortable. We had to do a combination of back riveting and normal riveting depending on access to the head.


I was not able to get the screw to work very good for me. We bought cleaveland tool spring back dimple die set. You can buy them from cleaveland or skygeek I believe. Might be overkill, but it makes a very nice dimple.

http://www.cleavelandtool.com/custompag ... -dies.html

We tried some cheaper version and it did not do as nice as the spring back version.

Mike

[Bryan Cotton]

I bought a prebuilt spar, but I have set lots of rivets. To set countersink depth when the rivet goes into an extrusion, practice on scrap. Set your countersink depth, insert a rivet, and check for flushness with your finger. You don't need the micrometer. When you are countersinking for a dimple, dimple a piece of scrap of the same thickness. Then countersink another piece of scrap and see how the dimpled scrap nests. It is cool that your Kung-Fu with measuring stuff is strong enough to measure a countersink, but a typical sheet metal mechanic would never do that.

One other thing, I would find/borrow/buy a bigger gun for the AD5 rivets. At least a 4X, and a 5X would be awesome. I got mine on eBay. And, if I was setting those rivets with a gun and bucking bar, I would not back rivet them. Pound the factory heads with a flush set and form the shop head with the massive bucking bar. That will tend to pound the heads in flush.

I've back riveted skins and light duty stuff. I have worked for two airframers and have never seen them back rivet the big iron.

[scotttro]

Thanks for the replies. I suspect I'm overthinking this whole process. What is confounding me, I think, is that there isn't really any way to confirm a good or bad rivet, other than cutting test blanks in half. It's good to know the test blank is good, but what about the work piece? It's a conundrum.

Bryan, what's the problem with back-riveting? My thinking was that since the "Sonex method" was basically that, I'd mimic them as closely as possible, albeit with a gun rather than a hammer. With all the practice rivets I've driven, I've gotten pretty good at the process, but I've got a flush set in the box if it's a better approach.

[wlarson861]

I back riveted my spars but I used a C-Frame that I bought from Avery tools years ago. Bryan is right about the size of the gun, I used a 3X gun to set the rivets but on every set I wished I had a bigger gun. with the 3x it has to hit the rivet so many times they start to work harden before they are fully set. If you can borrow a 5x gun great but a 4x will be better than what I used. As far as knowing the rivet is set properly it is based on measurements done before and after the rivet is set. The unset rivet should protrude though the part 1.5 times its diameter(5/32 rivet= 15/64) and after it is set should be 1/2 diameter high and 1 1/2 diameters wide,(5/64 by 15/64). I bought a set of "go-no go gauges from Avery tools years ago but you can make your own since the spars only require 5/32 rivets. If after setting a rivet you have the expected measurements then the rivet is good and will meet speck. Understand that these structures are way over built and Jeremy told me you can probably screw up 10% and still be strong enough.

[NWade]

Thanks for the replies. I suspect I'm overthinking this whole process. What is confounding me, I think, is that there isn't really any way to confirm a good or bad rivet, other than cutting test blanks in half. It's good to know the test blank is good, but what about the work piece? It's a conundrum.

Bryan, what's the problem with back-riveting? My thinking was that since the "Sonex method" was basically that, I'd mimic them as closely as possible, albeit with a gun rather than a hammer. With all the practice rivets I've driven, I've gotten pretty good at the process, but I've got a flush set in the box if it's a better approach.

I used a combination of back-riveting, regular riveting (with my wife), and the Sonex "big bolt" method. Regular rivet-driving was the best (but takes a good partner). Back-riveting was the least-consistent solution. For one-person operations the big bolt method was best, once I got everything set up with some blocks to hold things in alignment (in conjunction with the spar spacer tools that Sonex ships with the kit). I used the Sonex bucking-bar with both methods and highly recommend using it instead of something else. I suspect part of my issues with back-riveting had to do with the fact that I was using a 3x gun and I do think the rivets work-hardened some. The gun could bounce around a bit (despite working on my solid garage floor) and mash the rivet tails to one side sometimes; whereas a few whacks with the bolt lined up vertical would set the rivet pretty nicely on a consistent basis. The only downside to the big bolt method is that the bolt eventually yields and you need to make up a new one. I think I went through 3 or 4 big bolts, and another 3 or 4 smaller bolts who's heads I'd ground down to a narrow width to fit between some of the root-end stiffeners.

They're a bit buried since Google screwed up my photo albums, but you can see lots of photos of my setup and riveting progress about 2/3rds of the way down in this album:
https://picasaweb.google.com/1039641523 ... redirect=1

Good luck,

--Noel
Sonex #1339

[Bryan Cotton]

Back riveting is a great technique. It is used often for flush -3 rivets on skins. Consider what is going on though- the gun is trying to knock the rivet out of the hole. The bigger rivets need bigger forces to set them. This is where the immovable flat concrete floor comes in. If you are using a gun and bucking bar, and you front rivet, every whack drives the rivet into the work. The process works even without an immovable floor.

Is the rivet good? This part is easy, no cutting required. Is the factory (think rivet factory) head flush against the work? No smileys? That side is good. And ideal rivet starts off 1.5 diameters longer than the hole (proud, in tin-knocker parlance). After riveting it's height should only be .5D and it should pancake out to 1.5D. Don't get out the micrometer for this. When I stick a rivet through I hold an unset rivet up to it. That measures 1D. I estimate .5 on top of that. You can make or buy a gauge, and you can use a ruler too. I shoot for perfection but am happy with anything within two or three 128's of an inch (see stick hole thread). Also the shop head (think your shop) should be nice and roundish and centeredish over the hole. If you nail the rivet tail over it will generally end up oval. It is actually easier to set the big rivets as they don't tend to nail over.

One last thing- rivets and scrap are cheap. Set a bunch and get your skills and confidence. Then do real parts. There is somebody near you who is an expert. Find an EAA chapter. They will help.

[n982sx]

Here is a link to my logs on back riveting my spars with a 3X gun.

http://n982sx.com/sonexlog42/index.html

The 3x gun will work fine for back riveting flush or domed head rivets using the Sonex bucking bar on a table with a few caveats. I placed the Sonex bucking bar on a steel plate on the table and my table was very solidly built. Mostly I moved the spar as I riveted so the plate and bucking bar were near or right over a 4x4 leg.

If you are new to riveting, I do not recommend trying to set -5 rivets conventionally. Setting -3 rivets conventionally is a lot easier than -5's. I just finished 20,000 rivets on my new project and I back riveted any place I could make it work, especially with larger rivets.

The first page of my link talks about getting better dimples with what I now consider to be a sketchy practice. Do yourself a favor and just get the correct dimple dies from someone like Cleaveland tools. They will work in a hand squeezer if you can borrow one from someone as well as a C-Frame.

All that aside. There are 20 or so rivets where it is easiest to just use the hammer and bolt method. I did those on the floor of the garage and found it so easy to do, that if I was making the spars again I would have no issue with doing the entire spar back riveting with the hammer and bolt.