I'm still messing with the flanging trick, which I've dubbed The Duble. (i.e., do-able) The neatest part is that you aren't limited to circles. Make up your optimum triangulated shape, zip it out with a router and then use the routing mask to cast a die using a high-strength metal-filled polyester resin. Tests with Bondo did pretty well but there are some limitations; you must figure the amount of stress in the corner; Bondo craps out pretty quickly with thicker stock. Then you must figure out how to press the whole thing at once. If you just try hammering it cocks and fractures the cast punch. But pressed into place it works great.
Lots tricks in keeping things accurate; I'm using #6 nails & 5/64" drills to make my tooling holes. Little trick on forming the base of the flange -- you gotta route a groove to act as a foundation; it fills with the casting compound. Without it, the flange feathers off to a thin edge and shatters in use.
Right now most guys don't bother to punch & flange their ribs & formers because it's too much work. Router with an SCR controller makes short work of aluminum, so long as you have a properly made pattern. All 24 ribs took less than five minutes to cut out, including the holes. (That's a stack about 3/8" high; router ran kinda slow but it will do up to 1/2"). Making a punch & die to produce a perfect flange takes about 15 minutes and the only 'machining' is to router a radius on the hole in the pattern. (That's another area of investigation. Bigger the hole, the deeper the required flange, according to my 'stiffness' tests. Quarter inch flange does fine up to about 1.5" diameter. Big holes, 5" and larger, need a full half-inch. Everything else does fine with 3/8ths)
<<Doesn't the polyester (Bondo) exotherm a lot too? >>
Yes! But the shrinkage is part of what makes it work. I can usually pop the formed punch out of the waxed 'die' while it is still in the near-plastic stage, allowing me to use an air blast to cool it off. The epoxy tends to start smoking & locks up long before it is rigid enough to push out of the form.
<< But I wonder if aluminum paste from AS&S and cheap polyester resin from the hardware store would work just as good... >>
Maybe. I'm still experimenting. Today's test was to take a Bondo die, grind away the curved part and REMOLD it using metal-filled epoxy, needing only a fraction of an ounce to do a 2" die. Haven't tested it (cutting wood, other chores) but I'm hoping it will not break up the way Bondo does when flanging .040.
<< Are you forming the whole punch out of the goo? Somehow I pictured just a veneer of thermoset over a wood armature...would save a lot of goo, and reduce the exotherm problems by reducing the mass of material.>>
For the small holes, yes. And for most of the radiused corners. I shot some sheet metal screws into the punch-plate to give the Bondo a bit more tooth. For big, round holes I screwed a plug into the hole-space, leaving about 1/4" all around. You can't leave the gap much narrower or you can't be sure of filling the curve. I prod it down with the mixing stick, trim it with my pocketknife when it goes plastic, and start pressing it out when it goes solid.
For small flanges -- anything up to about 1-1/2", I use sockets. All the way down to tiny 1/4" drive stuff. Just drill a suitable hole in a block of wood, line things up and give the socket a tap with a hammer.
Sockets come flanged both ways. 3/4" drive, the drive-end is usually smaller, fairs into the socket-body with a neat angle. Make the hole to match, find a bolt to fit the socket, turn her over and whale away.
Going the other direction, just use an extension you don't especially like as your driver. Or hit the socket itself. Most small sockets have lovely cambered curves instead of sharp angles, plus a nice chrome finish that doesn't mar the aluminum.
Do a panel -- something like one of the curved bulkhead pieces -- it always comes out not-flat. Various tricks to flatten it out... or none at all, if it's not going to be see. One trick is to use the sockets again. Bigger ones, so they'll fit down over the flange. Lay the part on a solid surface, flanges pointing up, use the socket to gently tap down the area around the flange. Masking tape on the edge of the socket prevents doing any damage.
Sounds like a hoo-rah solution but once the part is riveted in place, all nicely painted up, nobody gives a hoot how it got its shape.
> I'm making tooling these recent evenings, last night I finished the fluting tool, I'll post a picture of it soon, and drilled an aluminum ruler for a rivet pitch gauge. Next up are rudder templates, and lightening hole flange dies. How many, and what sizes do you recommend?
I donno. I make them as I need them... if I really need them. For the wings, it makes pretty good sense to make them as part of your form blocks. For the bulkhead sides I'd probably use the arbor press and an individual die. Then again, I might make them as rectangular openings, mebbe 1x2. The work is being done in the flanges. Bulkheads should be mostly empty space. Those enormous flat panels aren't doing anything but providing some tensile strength and adding weight. Since there isn't much of a load in tension across the bulkheads, you could get rid of most of it.
> I'm just about ready to slather Bondo onto my first trial flanging die, and I'm wondering about the radii. I used my smallest roundoff router bit (1/8R), and I'm wondering if I should go a size bigger. The tiny radius I have on there now means a pretty shallow flange, liable to trip in buckling.
Okay, STOP. Grab a pencil and make a quick sketch of a 90-degree corner. Inside your 90, sketch in a uniform radius... size doesn't really matter. Now LOOK at the drawing. The HOLE is represented by the sharp corner of the 90. Guesstimate the ARC and swing it over to your uniform radius... and make sure you remember it's going to spring back by as much as 10 degrees.
I'm guilty of leading you down the wrong path here because flanging dies with rounded edges work best on THICK stock.
Does your router bits include a 45-degree? Howzabout a dovetail routing bit (ie, 60 degree)?
I've found that any metal thinner than .032 flanges best using the sharp corner provided by the 45. The reason has to do with the amount of flange you need to get the required stiffness. The first sketch will show you the mathematical relationship between the arc and the resulting depth of the flange, and as you're about to discover you need at least a 3/8" radius (3/4" rounding bit) to give you enough depth... that is, to form a flange (taking into account its spring-back) that bends over far enough to give you the stiffness you're trying to achieve. To calculate the flange depth with the 45, just use trig. Depth should be something like .7 the base. (Springback reduces it from the .707, etc)
<![if !vml]>
<![endif]>I've used a
hardwood dowel for the centering portion of the male punch. The dowel
is glued and screwed into a close fitting blind hole, which is in
turn surrounded by a routed recess, as indicated above, to avoid a
feather edge on the <![if !vml]>
<![endif]>generic auto body
filler is used on this die.
Somewhere over the rainbow... or perhaps out in one of the sheds... I have a hydraulic press-frame I made to do ball joints using a bottle jack. This was back before I had a floor-type hydraulic press. So I'm pounding the shit out of the outer panel spar webs, forming the flanges, mostly on the right side, and June comes out to ask what all the pounding is about. Just me. But what if the builder lives in an apartment house, barracks or dormitory? And besides, my knees were giving me hell. So I kicked around under the bench, found a small bottle jack, some 1/2" HRS plate, an old table leg and a piece of angle iron two feet long. Less than 30 minutes later I had a very compact pressing frame. Slide the shear web through it, position the flanging blocks, sit the bottle jack on top of the sandwich and do what comes naturally. Nine strokes of the lever later I got a perfectly formed flanged. Did the four small diameter holes in about five minutes. I've made a new flanging die for the 3" holes, still waiting for the filler to harden. I'll have to redesign it (the press frame) so folks can make it out of 2x4's but it's just too handy to ignore.
The big bench vise is faster for ribs and the like, and with the Acme thread and a 24" cheater, it's probably about as strong as the little bottle jack (odd rating... 8.5 tons) Using soft copper, I did a crush-test with the vise shortly after I bought it and got something like 20,000 psi for the area of the copper slug (about .5 square inch). You don't get that for the entire area of the jaw, of course but it does a fine job squeezing flanges into .020 stock.