090930 Skyscraper 03

I have been working on Skyscraper 03 for more than 12 hours and finally I've got the scripts working, STL ready, and renderable. Pretty happy about the result to be honest. And hopefully the 3d-print will work perfectly fine for my 1:1000 model.

I will go through the process from here on.


Imported the mapped-points-drawings into AutoCAD which the acquirement of points have been illustrated in the previous post.


Started mapping up the highlighted points using dots and crosslines.


Here's the results for the 4 categories


I decided to use square panel for the dots. The facades of the tower are also laid out with 10m x 10m square grids


Array across the panels with the 4 block panels. Rotate them to give more variations on the facades


It seems that 10m x 10m will contain a lot of points. So I made the grid to 20m x 20m and the density of points is just about right


Import the dwg/dxf file into Rhino


Applied Voronoi-2D script on facade and started to create some floor plates.


The wireframe generated by Voronoi-2D is then imported to 3ds max for piping (i.e. making from lines/curves to solid/mesh)


Isolated the wireframe to reduce memory usage for Scripts


I have modifed the script that I wrote in the previous post. This script is more simple. What it does is considering
-> each cell as individual
-> offset curve inward
-> CurvethroughPolyline
-> clean up the scene
This script can be made much better in terms of the return and the input variables.


Then I did a simple PlanarSrf command by selecting all curves in the scene to create the cutout surfaces


I wrote another script for this particular move - Extrude each surface to its normal.
Each surface has their own normal of extrude. If you type in ExtrudeSrf and select all the surfaces, you might encounter a problem where some surfaces might extrude along the construction plane which is undesirable. So this script does:
-> enter extrude height value (preset Cap and Both sides prior to runscript)
-> considering each surface
-> extrude individually via its normal
-> clean up the scene
It runs quite a while and this is the result with a thickness for the "membrane"


This image shows the combination of the primary cell structure + the membranes


Playing with the shapes of the floor plates


And stack them up with a rotating configuration. The idea here is to create some double volume spaces.


Here is the overall model!


Import the whole model into 3dsmax for basic renderings


Rendering in process... and below are the resulting renders.









I like this render the most.

090930 More Sites for Voronoi Cell

I just want to sum up all the research for Voronoi Cell with several links. Perhaps you can add more in the comment box.

Background:
Wikipedia

Examples:
seen.by.spiegel - voronoi-cell-cluster by Oilvan Oesterle, Michael Knauss
voronoi-diagram by imcyborg
Responsive Volatility - Programme

RhinoScript/Grasshopper:
RhinoScript Wiki
un didi - 3d-voronoi-in-grasshopper by Dimitrie Stefanescu
Rhinoscripting by Kenfield Griffith, John Snavely
Kokugia - RvbScript
RhinoScript.org

090929 Voronoi Script Test

The Voronoi Script (a.k.a Point Set Reconstruction) from McNeel Wiki is a little bit too primitive. It restricted to only abounded by a box and the cells are either curves, polysurfaces or meshes. It's really hard to manipulate the generated cells. To achieve what I want, which the cell membranes as primary structure while having a curve opening, I started learning how to script in Rhino.

Scripting in Rhino is not new and getting more refine and powerful. I have gone through the basic introductory of scripting called vbscript for beginners downloaded from McNeel Wiki. Then I saw RhinoScript101 on the web as well. Spent the whole day in trying to create openings for voroni cells. Something like this posted by sonniewong on Caustic Surface: 3D Voronoi Script Test.



Here's my result with the script written by me.

What I did was
-> Pick Polysurface
-> Scale <1> Extractwireframe
-> Join as Polyline
-> CurveThroughPolyline
-> PlanarSrf the Curve
-> End (that was as far as I got last night)





Ideally, the close polyline of the cells can for planarsrf with the insetted curve.
Then, I can extrude surface with a thickness. Let's see how it goes tomorrow.

090928 Voronoi Pattern Study

I challenge the skin structure of the towers by making it more organic with Voronoi Cell System. Before heading into the actual structure itself, I studied on what Voronoi Cell is about and precedented works. As described in Wikipedia, a Voronoi

"... diagram is a special kind of decomposition of a metric space determined by the distances to a specified discrete set of objects in the space."



Por Jorge Huang Li, the editor of VANGUARQ beyong architecture, blogged a post on "Futuristic cellular structure architecture". Basically he researched on the cellular structure in the mainly the field of architecture. Be interested to look at some real and virtual works from his blog.

So, here's some of the models that I have done with the Voronoi Script. It takes me a while to know how exactly this plug-in works. And the pipe command in Rhino just too inconvenient because you've got to do it on every curve. So I have to use 3ds max's modifier to give the spline a thickness.



-> editable spline
-> enable in view port -> radial (set thickness)
-> editable mesh
-> import back to Rhino (pain in the neck ... )

Here are a couple of results that I have made for the skyscrapers:


Learning how to use the PointSet plugin


Computed Voronoi 3D Cells


Voronoi Cells in rectangular tower


Voronoi Cells in circular tower

090926 Sunlight Study 01

I have done the first sunlight study on the proposed site. From my preliminary sketches in previous post, I proposed two towers on the east and west of the site to create a grand central axis connecting the Perth Concert Hall to the river.

The sunlight study is essential to see how high the towers can be and how the towers will impact on the site and the surrounding. Major issues would possibly be Views and Shadows.

In the first run, I have set the west tower at 300m high and the east tower at 220m high to see how much shadow they will cast around the site.









The date for lowest sun angle (winter) and highest sun angle (summer) is from the Sun Angle Calculator in Vray Rhino. It is more accurate than the 3ds max one.



In Perth, the lowest winter sun angle is on 2009-06-22 with Azimuth of 5.22113' and Altitude of 34.3702'. The highest sun angle is on 2009-12-17 with Azimuth of 18.9663' and 80.8911'. Here are the first set of results:

In Winter:




In Summer:




In winter, two towers will not cast shadow over surrounding buildings but not for the summer sun. The results show that there will be sometime that the Tower 1 (300m) will cast shadow on 1) east building of Swan Bell Tower 2) Tower 2 (220m) and 3) east building of Perth Concert Hall.

I guess it is possible to have perfect shadow if I flip the height of the two towers the other way round, i.e. Tower 2 is 300m and tower 1 is 220m.

From the sunlight study, the height of the towers can be determined. So hopefully, the result from second sunlight is positive which will be tested tomorrow.

090921 Skyscraper 02 with Attractor Scripted Facade Structure

The previous post has mentioned on the finding of the attractor script. This post is to illustrate how the attractor parameters being used to executing on to a simple boxy building. I hope the following print screens can help you in following my process. This building is a prototype exploring the ways in treating the fenestration.


Testing the RhinoScript see if it's working on this template


Draw a rectangle


Extrude to a desirable height


Rebuild the model to give more quad surfaces


Thinking of ways to explode the rebuilt-surface
(extract points, PlanarSrf, extract mesh, etc.)


Finally found a way, ExtractWireframe!


Wireframe structure of the rebuilt-surfaces


Splitting the surface with the wireframe structure


Creating some random points on the facade


Run the script


Computing the scale of element in progress


Final result with threshold <1> and scale factor <1>
This formed the fenestration of the building
Now have to cutout the holes.


Overlay the fenestrations with the original surfaces


Just in case they don't overlay on the same plane,
I move the fenestrations outwardly,
then do "Project" on surface to ensure the curves stamp on surface for "split"


Splitting the facade with the fenestration curves
Cleaning up the model


Extrude the surfaces to give a thickness


Adding solid frame to join all extruded surfaces together


Finally I do a 800 x 600 with V-ray
and this is the result

Time elapse: 4 hours (Modeling, Rendering, Blogging)