Posts Tagged ‘OMA’

x4_connected steel beams

As analyzed in the x2 the CCTV consists in a quite complex structure. In this case I tried to understand the connection of the different beams – vertical columns, horizontal perimeter edge beams and diagonal members. As the load on the structure changes from one point to another, there are several sizes of the columns, of which the smallest is 110 x 100cm.

different sizes of the columns and their location

There are many layers for the whole structure which are shown in the following drawings and images which I took – as the other illustrations – from the book CCTV by OMA, A+U, 2005

To allow the joint to connect beams comming from different directions, there is a additional element which takes over this function. Two connecting plates, which are also made of steel, are attached to the colomn and allow to fix the other structural members squeezed between them as shown in the drawings. To protect the steel structure in case of fire it is completely covered with a fire protection layer and a top coat.

bracing connection of structural members

my interpretation of the joint

The three main elements have different functions but only work together:

  • the vertical columns bring the forces directly down to the foundation.
  • the diagrid helps descend the forces but it’s main role is the bracing of the building – it is the part which gives the building its stiffness.
  • last but not least there are the perimeter edge beams which follow the edges of the concrete slabs and create the connection to them

This joint allows the connection of different inclinated elements. But all of them have to be nearly coplanar that they can fit between the two connecting steel planes. This fact limits the use of the element to places where the beams come together in a more or less flat surface.

x2_CCTV from OMA


To fully understand the following post you can read the last one which introduces the main ideas. X1_CCTVfrom OMA
As a source I used the book CCTV by OMA, A+U, 2005


THE STRUCTURE
The main structure of the CCTV is a continuous grid of diagonal steel beams – called structural diagrid – which cover the whole building. Where the loads are too big, the diagrid is doubled or even quadrubled. In addition to this structure there is a orthogonal structure which consists of vertical load-bearing columns and horizontal perimeter edge beams. These two grids penetrate the concrete slab in a certain distance from the façade. The diagrid is repeated on the outside where it holds the windows in place – this is actually what we see from the outside.


THE 3D MODEL
Taking the interpretation from the last exercice as base, I created a parametrical 3D model in Grasshopper which generates an abstraction of what the CCTV is.
As you move the z- or x-spacing in the model, the diagrid changes its proportions. The width and thickness of both diagrid and reinforcement can be set up individually. Until now the vertical columns and the perimeter edge beam are not really adjustable with parameters but the slabs change their position regarding the z-spacing – one z-spacing contains three floors.
The two intersecting planes generate automatically plan and section and they can be moved either along the z or the x axis. These line drawings are moved out of the 3D model and create a composite drawing with elevation, plan and section as you can see below. Next to this drawing a series of axonometric projections shows the simplified layering of the model.

make2D from baked GH-model

klick on the giff file showing the different layers


THE GH-CODE
download Rhino file
download GH code

To get into the third dimension taking the last GH-file was not really possible, so I started the definition from scratch. The new definition doesn’t build up a line grid in the beginning but it creates a series of rhombi which then are offseted and extruded into volumes. The functions for the attractors remain basically the same but the reinforcement is again a new definition creating another rhombus around each influenced point.
The parametrical model so far is vertical, but I think it won’t be very tricky to move this onto an inclined plane and using its u/v coordinates to define the volumes. Another problem to solve is the overlapping from reinforcement and structure. I tried several ways to avoid it but all of them faild, one is still desactivated in the GH file.
I was also thinking about using this kind of rhombi to apply on whatever surface inspired from David Fanos tutorial. But his base elements stay 2-dimensional and always just share an edge instead of a surface when it would be 3D (offseted).

overview of the GH canvas with explanations

close-up_main structure is generated

close-up_attractors are being set up

close-up_reinforcementis generated around influenced points

close-up_slab and orthogonal grid are generaded

close-up_plan and section are generated

x1_CCTV from OMA

sources: inspiration from the video tutorial by David Fano for the point attractors

image showing the forces in the construction

my interpretation of a given grid (black) and reinforcement (red) where needed

Following my interpretation the CCTV has a given pattern of crossing beams which are doubled where needed to reinforce the structure as analysed in the first image.
The following GH output is a result of attractor points and lines which double the structure in its influence radius/distance. The attractors are actually set in Rhino and can be moved to any place. Their influence radius/distances can be changed in GH. The grid from which I start is controlled by different sliders for x-spacing, y-spacing and the number of intersections.

composite drawing

flowchart for the GH code

overview of the GH canvas with some explanations

clos-up of the GH canvas where the grid is generated and attractors are set up

close-up of the GH canvas where the influenced parts of the structure are doubled

close-up of the GH canvas where the section and plan are generated

download rhino file
download GH code

Finally this code allows to set up different kind of grids and then reinforce them where needed. These locations can be defined by points or any kind of curves placed and moved in Rhino. Depending on the amount of reinforcement one desires, the influence radius of the attractors can be increaded or decreased with sliders in GH. Of course this is not only possible/useful for structural reasons, but can also be applied for other reasons such as shading.

x1_CCTV from OMA

sources: video tutorial by David Fano for the attractors

image showing the forces in the construction

my interpretation of a given grid (black) and reinforcement (red) where needed

Following my interpretation the CCTV has a given pattern of crossing beams which are doubled where needed to reinforce the structure as analysed in the first image.
The following GH output is a result of an attractor which doubles the structure in its influence radius. The attractor is actually a point set in rhino which can be moved to any place and its influence radius can be changed in Grasshopper. The grid from which i start can be set up in Grasshopper using different sliders for x-spacing, y-spacing and the number of intersection.

overview of the GH file

For a closer look at the code you can copy it from here and just set up a point within rhino or download my rhino file

attractor influencing the given pattern in its influence radius

To go on with this I’d like to try setting up multiple attractors or even a line which functions as attractor. But so far I couldn’t manage either of these ideas.