Does anyone know a good method to un- or reverse- develop plates? Rhino 5, Rhino 6

unhandled

#1

Hi everybody,

I am curious if anyone knows a good and efficient to recompose a 3D model from 5 developed plates.

While I only have the developed flat panels (single curvature, not compound), I’d like to turn them into a 3D model of the hull.

Basically I have a simple developable boat hull which is made up by 5 panels (=2x bottom plate, 2x side plate, flat transom.
The bottom panels joined along the keel and to the transom, the two side panels to be connected at bow, along lower panel side and at the transom.

Is there anyone who has done this before and might be willing to share their procedure?
Is there something like “roll developed surface” or “join edges 3D” and does this by “folding up” the panels?

Thanks so much.

Pretty


How to connect to surfaces along the edge?
#2

Hi Pretty, do you have the physical boat to work with? If so, you could find their place in space with curves and then try to bend your surfaces into the curves you made. Just a thought. —-Mark


#3

Nope.
Actual boat or 3D boat is not available.
All I have are the developed flat panels.

I’d like to do it in Rhino…

Thanks anyway.


#4

Just thinking out loud… Could this be done with Kangaroo… convert surfaces to meshes and then pull the vertices along the mating, naked edges together. I presume the panels are symetrical so the vertices should match up? I’m not sure if you can restrict the vertices on the rest of the surfaces to only move as if they were rolling up in one direction though.

If the surfaces are strips then they should convert nicely into meshes.

I think this is going to be tricky though as there is possibly many solutions if the final hull is not 100% rigid but maybe you can get close.

You might need to do 2 strips at a time.


#5

This will be tricky to get 100% correct. But if all you want to do is create some renders then it should be possible to manually iterate to a reasonable solution.

  1. Model what you think it looks like.
  2. Unroll the model
  3. Compare to the real cut files.
  4. Repeat until you are happy with the correlation.

#6

Unfortunately its not only about renders this time.
Needs to be correct to a fair degree. Boat is 3200mm long. Accuracy should be 5mm.


#7

Well this will be tricky because there are a lot of possible solutions. Thinking about it if you were doing it manually, actually building the boat to full scale or with paper/cardboard on your desk, once you join all the seams (with weld, stitching or tape) it will still be possible to distort the hull shape. The sides will be able to open out as the bow/stern rotate up. It isn’t rigid with just those 5 panels. Adding a midship frame will greatly reduce the possible solutions, but there is still flexibility. A gunwale part would lock things in well.


(David Cockey) #8

Pragmatic method:

  1. Print drawings of the panels
  2. Cut card stock, thin plywood or sheet metal to the panel shapes using the printed drawings.
  3. Bend and tape the panels together.

#9

Thanks David,

but this is not what I need. In real life thats easy, but in Rhino3D?
I am after an efficient, accurate method to do this within Rhino.

Ncik, your approach with the ruling lines is good, I was thinking about this, but its unfortunately far from a reversal of the unroll, flatten or smash tools.

Anyone else.


#10

Like David and Ncik already said, there are not enough constraints to derive a single solution. Try it with cardboard and you’ll see. If you can post your planes and you may get the best one can make out of it…


(David Cockey) #11

I didn’t say there is not a single solution. Prettypicturegirl said she has two bottom panels, two side panels and a flat transom. I assume one end of each side is joined at the stem. In that situation my guess is if the panel shapes are developable, ie no stretching, shrinking or shearing, then there is either one unique solution for the boat shape or no solution. (Other combinations of panel shapes could have multiple solutions.) However, and this is an important however, depending on the shape it may be possible to deform a physical version of the developable shapes relatively easily so that the panel shapes are not exactly developable.


#12

David is right in my point of view.

The panels are flat developped without distortion or compound curvature.
The length of matching edge curves are equal.
So in my eyes there would be only one solution as long as the keel to transom angle is known too, which it is.

Unfortunately I can not disclose the plans.

In principle the tool I wish there would be, is a match edges in 3D. So you chose the coresponding curves and the panels roll up until they match as good as possible.
Than you add the next panel with a matching curvature seam and so on.

I know that there are geometric constructions based on ruling lines which would perhaps allow this. Still that would be not what I’d consider an efficient method to achieve the result of a reasonably precise 3D model assembled from the formally developed curves.

So, if there is no such tool in Rhino (or a coressponding plugin) it might be food for the Rhino development guys to implement it eventually. I am sure that this is something which has its applications.

Thanks for your thoughts guys,

Pretty


#13

Let me know when that solution comes out!


#14

At least you mentioned card stock, so… check it out and see :wink:


#15

There are many solutions to this problem, so a computer isn’t going to be able to solve it easily. Build it on your desk with paper and taped seams and you will be able to warp it and bend it easily…ie, more than one solution.

If you have a gunwale part or a keel part/profile, then it might be solvable, but I don’t know how a computer could do that. Maybe that’s where kangaroo for grasshopper might have some success.

Alternatively, I think there is a garment development program, for real and cgi clothes, that can reverse flat patterns into clothes. But again, clothes can bend to a shape very easily, so there is no single solution there either.

I am building one of our kits on the desk right now to convince myself I’m not making a mistake. Will try to post photos when I’ve got something.


#16


#17

Well there is the mirror plane and with kangaroo it should be possible to deform the bottom plate so the edge is on that plane and another edge to the transom. Then probably three edges of the side panel can be bend towards the bottom panel, the transom and the mirror plane and minimizing the internal bending forces. The result should not be too far from the design intention.


#18

While I’ve designed numerous developable surface small boats, most of my work has been curvaceous. I just spent a short amount of time creating a four plane small boat: two sides, a bottom and a transom. I kept the surfaces as simple as possible. I smashed one side with the hope that if I moved the bow of that new surface to the appropriate location on the bottom panel, I would be able to then “bend” it around the bottom edge by merging control points. Alas, the control point are different for the smashed surface as vs. the original or bottom surface. Conclusion: once the Genie has been removed from the bottle there is no possible course of action to put him/her back in.


#19

You have hit the nail on the head for this thread Robb, the Genie in the Bottle analogy is very apt.

What would you actually call such a Tool, a De Roller? Re-Roller?


#20

a futile attempt at the impossible