Got an idea to improve the design while commuting. Change log from
previous version:
- Extra row of feathers (coverts)
- Curved the secondary feathers' terminal edge
- Slots on primary feathers are now monotonically decreasing in size
Here you can see the comparison:
|
Before vs after |
Design
The improvement started from the desire to add extra row of feathers on top of the secondary feathers, whose anatomically correct name is "coverts". Having this will reduce the blandness of the initial design.
My idea is to compress the transition units all the way to the top, which will free some paper and make the secondary feather longer. Then we can do a sloped pleat to create the coverts.
Here is the crease pattern of v1.0. We shall focus on the highlighted area to create the coverts,
I started again with this configuration. All the highlighted 1 unit flap just need to be distributed into 4 smaller pleats and spaced uniformly. This can be done more easily if all pleats start with the same shape. We could just "drop" it.
|
Before and after the drop. |
Distributing 1 unit flap to 4 smaller pleats is easy. Start with distributing it into 2 smaller pleats with gap in between. This is done by squash, spread sink, and the slide.
|
Splitting 1 unit pleat into two 1/2 unit pleats separated by 1/2 unit gap. |
Each of them will be split again into 2 smaller pleats, with gap in between. Normally we can just squash each flap, and closed sink it. It works on the right side.
|
A simple squash, and closed sink on the right half. |
However it doesn't on the left side because the pocket is not big enough for closed sink. There will be self intersection. There's an easy solution to just valley fold in half, unsink the inner layer, and do a closed sink here.
|
From top to bottom: valley fold, unsink, and closed sink the right half. |
The folding process is sequential, so don't mind the weird angled lines on the crease pattern. Now repeat it 4 times and we have successfully reconstructed the pleats for secondary feather.
I did a test fold. The sloped pleat didn't work but I found different way to do it. It's by lifting up one sunk edge, and now it can be valley folded with slope. Folding with real paper really does give idea.
The pleats are locked and hard to shape. It can be freed by pulling it apart, and sink in the paper like Ancient Dragon's legs. This caused the pleats to behave as if they are made from the paper's edge, and shaping them would be easy.
|
Repeat for the remaining 31 pleats 💀
|
|
How it looks like in crease pattern |
I did a new test fold because that green double tissue was from an old batch when I was bad at treating paper. The tissues are splitting apart on various parts which is disgusting to fold. In this new test fold, I found an issue that also exist on the previous version. There is a gap between secondary feather and primary feather. Closing this gap with mush is possible but I would prefer not to. The structural way to fix it is by shifting the pleats so they are closer to the primary feather. Should be as easy as mirroring the transition unit...
|
An annoying gap |
|
Mirroring the transition will shift the pleat |
... or so I thought. Upon closer inspection, the shape of the pleats to be distributed is not all symmetrical. On the long and skewed diamond, you can see that the left side has ridges that go straight while the right side's ridge is bouncing. Asymmetry like this means we can't simply flip the previously found transition.
|
Ridges highlighted. Notice that two with diamond shape is symmetric, but not the other two. |
So a different transition unit is needed. While it is not hard to figure out, it is a shame that not all transitions are the same anymore.
|
From top to bottom: squash symmetrically, slide the left half with squash fold, squash+closed sink on the left half, and finish the other side just like the initial transition unit.
|
Hold on, couldn't we just mirror everything? I tried, and the result is either we get more smaller layers around the 22.5, or the pleats are too close to the primary feather.
|
Totally mirrored transition for secondary feather pleat, leaving a space that is too small for the remaining primary feather. See bottom left part.
|
I made another test fold and it worked. There is an additional benefit that the unsink near tail feather is not crossing any pleats anymore.
|
The circled part is no longer crossing any pleat. |
|
The serious test fold |
|
Cross section of the wing |
There is an extra improvement by curving the edge of the secondary feathers. This is done just by valley folding the edge with slight direction change on every pleats. It was too flat on the previous version and I think it looks much better now. It is one of those little things that human prefers, like drawing curved clothes is more interesting than straight clothes.
Another improvement is on the primary feather. Looking back on v1, the slots between each primary feathers go like this: big, big, big, small, big. It's not pleasing, and it should have been monotonically decreasing from big to small. So in this version I allocated more paper on the 4th slot. This makes the wing much better.
|
Making the the slot between 4th primary feather and 5th to be larger. Not going to draw the full creases as it would be too complex. |
I am satisfied with all those changes. There's an O-gami paper I bought specially for this model. Given the price ($15) I don't want to use it straight away. So I made yet another test fold with double tissue. This test fold would be as serious as if it is final fold. In the end it is still beneficial as I could play around the final posing.
This time I chose landing pose. The last version's final fold on soaring pose is poor at showcasing all features but the primary feather and head. So all the effort to shape the legs, tail, and wing surface are wasted. Landing pose "flexes" all features so I hope it can express them well.
|
The folds from the beginning |
Now I am fully satisfied with it. At least for now.
My O-gami is the thin one. It has non-uniform thickness. At some area, it felt like single tissue, and in others 1.5x single tissue. In average it is about 1.2x single tissue. This paper has excellent crispness and crease reversal. It just snaps to the other direction perfectly.
I had to summon a lot of motivation to finish the fold. It is quite tedious to fold all those feathers and toes. Due to skill issue, some parts look wrinkled.
|
Fresh O-gami paper |
|
Initial square |
|
Precreasing... |
|
Arrive at this form to continue folding sequentially |
|
Half collapsed |
|
Initial pleats created |
|
Pleats fully distributed |
|
Finish collapsing |
|
Legs done |
|
Freeing the trapped layers on the covert |
|
Showcasing the wing structure |
|
Before final shaping |
Comments
Post a Comment