The Big Bend (Jan. 18th 2005)

It's for sure now the brake will not work as it currently is. So I disassembled the machine again and changed the critical parts:

Watch the car jack in the middle of the plank pushing on the ceiling. Without this additional push the bend would become uneven.

However great care is adviced when doing dangerous things like this. There was tremendous pressure on the jack (I estimate up to 1 ton). If the plank flips to one side because the push is a bit assymetric then you're better not nearby. This assembly can explode! For safety I attached this horizontal post which takes some of the side load.

 

 

Beginning of the bend

 

another view how I safetied the jack. Also the beam to the ceiling is tied up so it will dangle around and not damage the wingskin if the construction crashes.

 

 

The final result. Unfortunately the bend is not the expected 85 degrees but close to 100 degrees. I'm totally uncertain if the skin can be used this way. Does it make sense starting to drill the pilot holes?

 

What I also have learned:

This is the 1:1 leading edge template fom the Sonex plans. This (horizontal dot-dash) 'Leading Edge Ref Line', whatever this should be - I don't know - definitely it is NOT the mean aerodynamic chord and it is NOT the tangent where the bending-press tube touches the sheet metal !! See the circle I drew inside the leading edge. The center of the circle (and the 'real' bend reference line) is 4.8mm above this ominous 'Leading Edge Ref Line'.

=====================================================================

 

Leading Edge Press Problems (and a Solution in Sight) (Jan. 15th 2005)

At both ends of the leading-edge-press now are two pillars. These transfer the immense bending force to the ceiling. Even that way ist was still impossible to do my first test-bend. The force which builds up by the threaded rods could not be applied at the center of the press because the solid vertical plank bowed up. Only by the help of another car jack positioned in the center of the press the bend could be accomplished.

The bent skin is very hard to remove from the press (even that my test piece is only half the width of the real wingskin). The full wingskin can only be removed by disassembling the bending brake.

Seems this bending job develops to a real problem.

Here is a cut off piece of the testskin layed on the Sonex plan. It's obvious that the bending radius of the test piece is way off the nominal shape.

What's going on here?

Currently my opinion is that the Tony-Spicer-style bending press cannot work as expected when used with a 47,6mm outside diameter tube. The 6061-T6 aluminum has (actually all alloys have) a 'spring-back' habit. In other words: after you bend the aluminium around a round bar and let loose, the aluminium will spring back and the resultant bent piece will now show a much larger diameter than the bar it was bent around. This is 1st grade metalworker's knowledge.

After a first eyeballing the springback effect is ~a lot~. I will do some testbends, starting with half of the nominal diameter and progressing until the (almost) ideal shape is found.

I wonder how all you buddy builders have done it (maybe you should all check your actual leading edge radius...)

 

...some hours and several bends later

This is the 'experimental' test brake. I started with 25mm diameter (much to small) and ended for today with 33,8mm (an old zinc-coated water supply tube). This diameter is promising.

 

 

This looks about right (33,8mm tube used). In the past I often heared about tremendous pressure on the first row of clecos when fitting the leading edge skin. Could this too-large bending radius be the cause?

 

from left ro right:

 

(1) first bend, wrong tube used, wrong radius, wrong shape

(2) 33.8mm tube used, encolsed angle too small, this leads to an unnatural circular nose and too straight tangents

(3) 33.8mm tube used, wider angle however a separation line where the circular section starts is still visible. The shape should more 'flow' ellipsoid-like

(4) angle still more open, provides the best shape but requires high bending force on the wingrib.

=====================================================================

 

Skinning the Wing (Dec. 23rd to Jan. 10th, 2005)

This is the left aft wingskin, already pilot-drilled. The mainspar forward flange is levelled, so the long row of rivet holes can be drilled . I used this tabetop drillpress for the upper row. The lower row I drilled free-handed using the airdrill. Proper levelling of the wing structure is very important here, otherwise the skin-holes will not match the rib-flanges (and - what's worse, a warp will be built into the wing)

 

 

after another 5 hours later all rivet locations in skin and ribs are deburred and dimpled. I'll dimple ALL rivets holding the wingskins, - except - the forward-aft-joint rivet row which goes thru the mainspar. Because I don't like to weaken the mainspar by countersinking, protruding rivets will be used here. Deviating from the plans I use 0,032" / 0.8mm sheet thickness (instead of 0.025 / 0.6mm) per the plans. Hope this will reduce the buckling tendency when high torsional loads are on the wing.

 

 

This strap-belts allow an easy flip-over of the wing (you'd never believe how often this wing had to be turned over). Furthermore, by using the ratchets it's easy to wind the wing up under the ceiling (to free the table for the next wingskin)

 

 

this is the leading-edge bending brake à la Tony Spicer style. It was a hard piece of work during a whole weekend long to get this construction together. Instead of PVC conduit I use a steel tube. Because I never ever would drill anchor holes into my nice wooden shop floor, I received very disappointing results on my first test-bend. Both ends of the lower plank bend up about 5 cm on the right and left ends when I tightened the M12 nuts. I stopped when even the table (to which I clamped the press) started to lift off the legs at the far ends. It was totally impossible to do a full 85 degrees bend this way.

 

 

The next idea is to forget the threaded rods and use three hydraulic car-jacks instead. The jacks will work against pillars which stem themselves against the ceiling (my basement ceiling is made of 2 x 4 meter sections of steel-reinforced concrete. The weight of one of these sections is several tons, so this little push from below most probably will not lift the floor of my living room upwards.

=====================================================================