Weaverbird is a topological modeler that contains many of the known subdivision and transformation operators, readily usable by designers. Instead of doing the work repeatedly, or sometimes using complicated scripts, this plug-in reconstructs the shape, subdivides any mesh, even made by polylines, and helps preparing for fabrication.

Weaverbird – 0.7.50.0 Spring GH. For Rhino 4/5 and Grasshopper 0.9. In wb toolbar. Please upgrade to this version if you are using any earlier releases. This version will expire on December 31, 2014.

To install, right click, choose “save as…” and then double click the file to launch the setup. If you encounter any problem contact me with a comment.

If you don’t have admin rights (you will find instructions in the zip file).

Main topological commands:

Catmull-Clark smoothing (wbCatmullClark). Calculates the type of mesh-based recursive subdivision described by Edwin Catmull and Jim Clark, at first in 1978. The resulting mesh always consists of quadrilaterals.

Split mesh into Quads (wbSplitQuad). Calculates a new mesh, which is formed of only quads and generally appears similar to the old one, except that it is welded. It is topologically equivalent to the Catmull-Clark subdivision.

Loop smoothing (wbLoop). Calculates the type of mesh-based recursive subdivision described by Charles Loop, at first in his Mathematics thesis in 1987. The resulting mesh always consists of triangular faces.

Split mesh with inner face (wbSplitPolygons). Places a new face departing from the middle of each original face edge. Caps the remaining hole with Sierpinski triangles.

Sierpinsky Triangles subdivision (wbSierpinskyTriangle). Places a triangle in each corner of a mesh face. The mesh will have one more hole per face.

Frame (wbFrame). Computes a new mesh with higher naked edge count, where each face has a new hole in the center and resembles a picture frame. The resulting mesh always consists of quad faces.

Carpet (wbCarpet). Computes a new mesh with higher naked edge count, where each face has a new hole in the center. The resulting mesh always consists of quad faces, and can be used to compute a Sierpinski carpet.

Window (wbWindow). Replaces each original mesh face with a new one, reconstructed on the inside. Each face has the same number of sides as the original one.

New mesh primitives definitions:

wbPrism, wbAntiPrism, wbPyramid, wbDiPyramid

Additional commands:

- wbOptions. To see the main log, set up the threading model, control the maximum time a single operation is allowed to take, and set options for the Loop subdivision scheme (to use Loop’s receipt or Warren-Weimer’s one).

- wbProperties. Opens a window with topological information about any mesh object in Rhino. Exposes triangular and quad face counts, edge count, current and welded vertex counts.

This list of C# scripts is a work in progress and is open for requests. Supporting  v. 0.8.2 and some previous versions.

baking & Referencing
	bakeName GH 0.6.0012 GH 0.6.0018 GH 0.6.0043	obj (geom) – a list of objects to bake name (string) – a list of names (common Rhino property) activate (bool) – one go-no-go signal, or a list thereof A (strings) – a list explaining the result(s)
	bakeAttributes GH 0.6.0012 GH 0.6.0018 GH 0.6.43 GH 0.8.2 GH 0.8.7 GH 0.8.52 w/ groups	obj (geom) – a list of objects to bake name (string) – a list of names layer (string) – Rhino layer name(s) color (onColor) – screen+print color(s) lineT (string) – the name of a linetype. Please only use with custom names. pWidth (double) – printing thickness(es) wires (int) – number(s) setting the quantity of wires, as in the property panel material (int) – GH shader(s), containing a Rhino material activate (bool) – one go-no-go signal, or a list thereof A (strings) – a list explaining the result(s)
	findID findRef GH 0.6.0012 2009 06 19	x (geom) – a Rhino geometry A (string) – The Rhino ID A (MRhinoObjRef) – The Rhino Ref
Automatic Sections
	contours GH 0.6.0012 GH 0.6.0043 2009 06 04	input (geom) – the objects to contour axis (onLine) – a line identifying the origin and end of the sectioning axis dist (double) – the interval between sections geomOut (geom) – a list containing all geometries meshOut (geom) – a list containing all geometries deriving from a mesh brepOut (geom) – a list containing all geometries deriving from nurbs-types pts (geom) – a list containing all geometries deriving from curves and lines
Continous redrawing
	theEngine component for GH 0.6.0012 Compatibility release for GH 0.6.0018 available, but users are encouraged to use the new timer within GH. GH 0.6.0043 GH 0.8.0003 New 2009 06 06	isOn (bool) – one go-no-go signal speed (int) – a period, expressed in milliseconds, that sets the refreshing rate resetAt (int) – the last shown number in the series, which will be restarting at 0stopAt (int) – the last shown number in the series starting at 0 reset (object) – any object (or bool) passed here will make this component restart from 0 C (int) – the current count
	Buckets GH 0.6.0012 GH 0.6.0018 GH 0.6.0043 2009 06 10	x (list) – a list of points. For the bucket, it can be a list of anything. reset (object) – any object here will make this an empty list limit (OnInterval) – two numbers which set the max and min values for each internal point coordinate recov (int: 0, 1 or 2) – a method to recover in case the internal value exceeds limits: wrap world (restart from other side of domain), limit (stop at the threshold, or restart from middle of the interval. A (list) – the list in memory
	renderAnimation GH 0.6.0012 GH 0.6.0018 GH 0.6.0043 GH 0.6.0059 GH 0.7.0055 GH 0.8.0002 GH 0.8.0004 GH 0.8.0013 GH 0.8.0065 updated by Jacek 200906 10 (four examples, see last one)	geometry (list) – a list of objects to be added to the scene before rendering viewport (string) – the name of the viewport to use. If empty or non-existing, the active one will be used filePath (string) – a name of an existing folder. If empty, the folder of the open document will be used. If document not saved, the current user’s document folder will be used fileExt (string) – jpg, png, bmp, or another extension that the rendering engine supports material (shader) – a GH shader. If material has a name and name is already existing on the scene, then the material on the scene will simply be inherited. No need of double materials current (int) – the sequential number of the frame rectMem (bool, defaults false) – deletes redo+undo after each rendering activate (bool, defaults false) A (string) – a text, explaining the outcome
fonts
	txtLines GH 0.6.0012 GH 0.8.0061 New	face (string) – the name of the font bold (bool, default false) italics (bool, default false) size (double, default 5.0) content (string) – the text shown pt (on3dPoint) – the point ancoring the text A (OnLine list) – a list containing all curves
Networking
	updReceiver udpSender GH 0.6.0019 Processing 1.0.5 + UDP library 2009 08 31	port (integer) – a resonable port number out (string list) – network echo client (UdpClient) – a .net framework udp client A (bool) – message received B (string) – message

Last week I have been visiting my brother in Barcelona. Maybe I should blame the maritime environment if I started experimenting with knots and threads. So, what I am going to show you today is a method to digitally knit yarn in Grasshopper. A similar definition to make a chaining fabric is provided at the end of the tutorial as well.

The stockinette stitch pattern

Download this Grasshopper definition

We are going to model the common stockinette stitch pattern at first, because this stitch can be easily adjusted to other knitting, chaining or weaving fabric compositions.

1. The Grasshopper file starts from a straightforward definition of a knitting tile made out of curves, a surface and some parameters. This tile is illustrated in the picture above and is exactly the one marked in the dashed “component” box. The tile is surrounded by a rectangle in the grasshopper canvas, so that you can quickly adjust it to other definitions, in case you should need.
2. A containing box is created, departing from the input curves. We are particularly lucky this time, because the open part of the curves can be exactely used to define the boundary box.
3. This tile of curves is then repeated on the surface using the surface morphing component of Grasshopper, which is particularly useful because it maintains the geometrical continuity of what is being positioned.
4. All curves are joined and rebuit using a custom-written C# method.
5. The piping component finally produces the pipes.

Chaining pattern

Download the chain definition

By minimally modifying the previous definition, we can obtain a series of chains on a surface. It similarly works with one tile, which is then copied on the surface. We could expand this definition by making each newly formed link (a closed curve) periodic.

In cartoons and movies, animation is achieved by rapidly displaying a sequence of pictures that have a common structure. This method creates the illusion of movement, because our brains will, as they are used to, reconstruct the pattern of motion from the static pieces.

How many frames

As we are going to animate an object, we will make it follow a curve. We need, therefore, to sample this curve in a number of (small) intervals. The number of intervals will determine the number of final frames we produce. If you plan to show your animation, later, on a TV screen, then you can take the frame rates of the PAL or NTSC systems: the European PAL has 25 frames per second (fps) and the American NTSC 29,97. If you plan to show your video online, you will need relatively fewer images: at least 18 fps, but most times 24 or 25 fps.

• For a European minute, therefore, the computer will need to render (60 s * 25 frames/s = ) 1500 frames.
• Similarly, for an American minute we will need (60 s *  29,97 frames/s = ) 1789 frames.

Subdividing the curve

To make the movement better controllable, we can, instead of subdividing the curve by Euclidean distance, evaluate it at a fraction of its domain (in its own parameterization). We will therefore be able of applying a weight to a control point, and it will influence the speed in proximity of the respective knot.

• A higher weight coefficient “slows” down the movement in the area.
• Two or more control points close to each other will make the animation basically stop for some time.

Moving the objects

RhinoScript gives us a number of methods to define the position of objects. For example, we can calculate the volume centroid of closed Breps, or the area centroid of open meshes. We will use these points to define the original location of our objects (O).

Because we want to follow a curve, we will need to move the object to the curve beginning, or S. For each frame, we will move by a t parameter farther on the curve. To avoid small summing errors, the script takes again the original coordinate and then sums the vector OS and SN, and does NOT move every time from the previous location (P) to the next (N). This would be subject to errors for small numbers, or for many frames.

Download

• Draw a curve, then run animator.rvb first and activity.rvb at last.

Implementation

animator.rvb
The animation details like the ID of the object we move and the original position are saved in a class called animator. As the namespace for VBScript is shared, we can define the class “animator” only once. Therefore, you can load the animator once for every Rhinoceros session. The class does the job of finding out which type of object we are dealing with on its own, so we just need to feed it an ID after initialization with the method setUp(). When we destroy the object, such as in the case we stop the script by pressing ESC or we are finished with our script, the object original position will be restored.

activity.rvb
The activity file will create one sphere and look for the first instance of a curve in the document. Make sure that it can find it. Then, it will make the sphere follow the curve in 750 frames. The constant DontRenderUpTo will skip the rendering part of the loop for all 750 frames.

Other examples

Many other ways to achieve motion in Rhino are explained on the Grasshopper tools page, at the renderAnimation link.

How to use JBullet

To use Jbullet easily in the new processing 1.0, you can just follow these steps:

1. install Java3D to have access to some standard matrix operations in Processing & Java. What you really need is only the vecmath class. If Processing (depending on the version & settings) still does not see it, put the vecmath.jar inside processing’s libraries\vecmath\library.
2. download Jbullet and be sure that you put jbullet.jar inside processing’s libraries\jbullet\library.
3. restart Processing if it was running.

Download the books.pde source code. And please, if you find this useful, post a link to your sketch or video that you did with this, I’m curious!

Addendum: if you want to use P5sunflow, then use processing v135: the version 1.0 seems not to be working with it. The code of the sketch is compatible, but you have to install the libraries in the processing folder itself.

Video

A video is on Vimeo.

Librarian’s nightmare & dream from Giulio on Vimeo.

Web edition

This release can run in your web browser. It uses P3D as software-simulated three dimensional enviroment.

I suggest you try these commands: 1 (then grab a box with right mouse button), then 2 (try zoom with central wheel) and 3 (orbit by dragging outside of a box), 4, 5 (adjust the sizes of the volumes by dragging the top bar) and 6 (custom views).

If you have problems, try downloading the latest Java or watch the Flash video, instead. Built with Processing.

Processing sketch source code

You can also download the full-screen, desktop (Processing) version that runs quite faster thanks to OpenGL. Interactive Landscape Tool (zip).

Special thanks to Axel Kilian for the help during the design explorer class (beginning of the development), the whole Hyperbody Research Group in TU Delft for feedback and the Processing community.

Download: pdf slideshow – pptx presentation

I wrote myself, for example, to David Rutten, who has been a good source of understanding for Rhinoscripts. I thanked him for sharing his code.

NURBS programs like Rhino and Maya allow you to generate objects from many adjacent surfaces. This diagram shows how to work with continuity among these surfaces, and also curves and lines. Position, tangency, curvature radius and derivatives are used to graphically explain G0, G1 and G2 to every audience.

A wall is usually constructed as an entity that separates an interior from an exterior, leaving the exchange of air and light to the parts called windows. The motivation for this is notably the idea of shelter; that layer of separation between the human skin and the elements.

This little experiment wants to challenge this assumption and proposes a small component that is capable of sensing the presence of people. When someone is close to the wall, the partition become more porous: it allows the outside of the wall to see the interior.

At this time, light and air are allowed to come in. Additionally, this can constitute a manner of showing the presence of inner activities in a public building.

Behind the scenes

Visit adding motion in RhinoScript to learn how to put together a similar animation.

I am just back to Delft after having visited Barcelona to participate as a Rhino educator to McNeel’s RhinoScript workshop. Luis Fraguada of Live Architecture Network was the instructor and showed us some useful techniques for geometry construction and data visualization. I developed one of the script of the class a bit further and came up with this DNA helix.

Download the code: dna.rvb.

Drawings: #5 is made by Ryoko Ikeda, #9 by Ilaria Giardiello. You can submit your own code by simply filling the fields below here (use “make code”).

Source code: bezienne Bez Point controls copyPaste pointEditor save. Built with Processing.

Download the source code: sakura.rvb.

“Option Explicit” is a series of scripts developed in collaboration with Ryoko Ikeda that deals with organic, abstract form and contemplates experiments that have been dealt with in aspects like geometry, light and shadows.

The intermediate goal is computationally explore shape creation, using the implementation of VBScripting that is offered inside Rhinoceros 4.0. The more far-seeing goal is to investigate algorithmic relationships among entities that usually surround us and re-establish their creative articulations, to simply allow us to deal with complexity.

Many things in nature grow from little to big. In fact, many people, when asked to name things that shrink, cannot normally name any for some time. But how does a tree actually grow?

Lindermayer was also fascinated by this question and develped a whole set of explanations, as a part of wider sintax-free grammars, to explain it. L-systems explain the groth of trees through a simple, unfolding, enlargement process, that we can imagine as the restarting every spring.

“Just a small change in the script of The crown and we get closer to an autumnal view.”

Download the source code: tree1.rvb.

“Option Explicit” is a series of scripts developed in collaboration with Ryoko Ikeda that deals with organic, abstract form and contemplates experiments that have been dealt with in aspects like geometry, light and shadows.

The intermediate goal is computationally explore shape creation, using the implementation of VBScripting that is offered inside Rhinoceros 4.0. The more far-seeing goal is to investigate algorithmic relationships among entities that usually surround us and re-establish their creative articulations, to simply allow us to deal with complexity.

This experiment, on our side, has the goal of understanding fractal object as a way to obtain morphogenesis. Each generation translates, scales and doubles. How to make it really work with rotations? The best answer would be quaternions, but also rotation matrices would do.

The construction curves are showing how the growth path resembles patterns that are typical in
nature: similar to the one that biologist observe in corals and other sea life.

Download the source code: crown.rvb.

“Option Explicit” is a series of scripts developed in collaboration with Ryoko Ikeda that deals with organic, abstract form and contemplates experiments that have been dealt with in aspects like geometry, light and shadows.

The intermediate goal is computationally explore shape creation, using the implementation of VBScripting that is offered inside Rhinoceros 4.0. The more far-seeing goal is to investigate algorithmic relationships among entities that usually surround us and re-establish their creative articulations, to simply allow us to deal with complexity.