Origami Coconut Crab

Coconut crab is a species of hermit crab that has abandoned the lifestyle of carrying gastropod's shell. Instead, it grows hard exoskeleton and curl its abdomen underneath. This allows their growth to be unconstrained by gastropod's shell availability. As such they can get really big. Unfortunately their existence is on vulnerable state due to habitat loss and over hunting.

The Origami-Dan's design contest theme for January was "improvement". We can submit improved version of our previous design. I guess it's good thing to do to start the year, to look back and see how much we have improved.

So I improved my coconut crab design into version 2. Here's how the old version looked like, which shall be called version 1:

Ugly, I know. Moreover my phone was so bad that the picture it took isn't clear. But you know, I was really proud when I finished it. It was my first boxpleated model. The crease pattern:

Design

Looking at version 1, here are the list of flaws:

1. Head and appendages around it are too big relative to the body.
2. First pair of claws are too narrow.
3. First two pairs of legs are too short.
4. Not enough leg segmentation.
5. The back plate should be have the "OO" pattern.

Below is the reference picture I used to come up with those list.

Source: Dictionnaire d'Histoire Naturelle of 1849
Taken from Wikipedia.

Next I took some time to think what is the right design approach. This creature has many long appendages, four of which is forked. It made me think that boxpleating is still the right solution.

Solving the first and third flaw about head proportion and leg length is easy. Just alter the tree representation and we're good. The second flaw about claw's width should be somehow solvable using level shifter so I'm not worried about it. Fourth flaw about leg segmentation can be solved by adding extra length on the legs to accommodate for pleats. 

The fifth one wasn't easy for me. I was thinking to come up structure like this:

Somehow I need to transition from axial+2 and axial-2 pleats with seam into axial+1 and axial-1 without seam. I didn't know how to make it so I tried free folding. Start with usual accordion on 16 grid, but level shift some of the layers in the middle to make the seam.

Level shifter that is not in the ridge always consume a pleat, so I made a pleat.

Later I found the transition from seamed region to seamless region:

Open the seam on the other half.

Closed sink the axial+2 into axial+1 on first layer.

Repeat the same for the second layer, but this time we need to inside reverse fold the edge first.

Closed sink on second layer completed.

Repeat on other side, and transition completed.

Do this on the other side, and problem solved! What's needed is a middle flap with length of 1 unit to transition from head to back plate, then another similar flap to transition from back plate to abdomen.

With all flaws addressed, the packing can start. I surely remember how long did I take to come up with the packing for version 1. So this time I will try using Box Pleating Studio by Mu-Tsun Tsai. Hopefully more efficient packing can be found through pythagorean stretches. Below is the tree to represent the crab based on the reference picture above.

The figure above is actually the simplified tree. I didn't show that the claw and third leg should be forked, as well as the abdomen that should be pleated. This allowed me to drag and drop the squares more easily since there are less moving parts. Just keep in mind that the claws should have extra width or height.

With the aid of BP Studio, I came up with this packing quickly. Not too different from the packing on version 1.

Notice that lclaw, rclaw, lleg3, and rleg3 has extra space for the forked tip. I tried to fit everything in a pretty grid size, and was lucky that 48 fit well. The pythagorean stretch is a "perfect pythagorean stretch" because it forms the 8, 12, 15 triangle; so I felt lucky once more.

Then I used Orihime to draw the fully assigned mountain/valley crease. 

There is still one deferred issue: the claw's width. I tried to snug level shifter to widen it, but no success. Then I got an idea, what if I just spread the flap? Stupid idea but works.

If it sounds stupid but it works, then it ain't stupid.

Fold

I had a feeling that something will go wrong so I better test fold. There was a double tissue I purposely made for test fold because the tissues have defect. Collapsing was straightforward, although the sheer thickness made it somewhat annoying.

No way I can shape this thick base into the crab. Still, making test fold is beneficial since by that time I was assured the flap length and width works. The claw's width was good as well.

Wider claw.

The final fold would need single tissue. Kind of afraid since single tissue is so flimsy and hard to reverse crease. I had a bad experience when trying to fold KoSBoP with it.

This time I slather another layer of MC on top to make it extra crispy. I consulted friends in Origami-Dan for collapsing tips with single tissue and braced myself. With patience I finally overcame my fear.

Collapsed base.

Shaping took some time because of the layers. I regret that I didn't reinforce the back plate with thicker paper from underside (a technique popularized in Origami-Dan by Kimiro). The back plate is so thin and fragile. Fortunately other parts went well.

Patience is something I learned from folding this model. There's one part that I am not sure... which is the abdomen. I don't know if it should have used axial+2/-2 instead of axial+1/-1. Seems that axial+2/-2 would allow better shaping since the abdomen's edges can be curved downwards and creating more 3D shapes. To achieve that, just don't do the closed sink to narrow the body and do it only on the underside. The CP is shown below.

That's about it for this model's creation. Big thanks to Mu-Tsun for making BP Studio and make origami enthusiast's work more efficient.