Oct 5, 2009

Conclusion/Recommendations

So there it is.

A skin that can be interpreted at different scales. At a distance, it is a sculpture weaving out of the art precinct of James Street Mall and connecting with the shoppers in new Sarugaku. The users of the bridge can interpet it as a sculpture from the wood cylindrical skin, or as the path that they use to get from A to B.

Drivers going under the bridge can also interpret the bridge as being affected by the constant throng of traffic on Beaufort Street.

There are obviously other methods of determining how something is interpreted at different scales. I would classify the bridge as a suggestive method. Another interesting possibility that could have been explored as well is the use of LED technology, which has come in leaps and bounds.

For example, the Ilumina building in Singapore uses such technology and it would have been interesting to see how that would have worked at a scale such as the bridge. Or maybe even smaller like the bus stop that was never realised.

And extending that idea a little further is the use of LED technology that reacts when it is touched (see picture second from bottom) which could indicate to the one way traffic of Beaufort Street when people were inside the bridge.

My only regret is that I didn't have one of those weather forecasting supercomputers which would have been required to realise the perforated wooden cylinder. Anyway, it seems to work in section and elevation (and obviously in plan... it's just a path after all) so it's not all bad.







The failure

Because I worked so damn long on this one, only to have it fail on me. I thought that I'd at least provide a vague idea of how it should have work in perspective.

Damn you 378Mb file.

FINAL final renders

edited in photoshop, with background and context.

For your viewing pleasure.









Final renders

Rendered them all out...

now just have to edit them in photoshop to make them look decent.



Skin is skin...

Since I'm trying to create something that is interpreted as sculpture/path/structure at different scales, I guess any nice skin will work.

I'm borrowing a script from http://neoarchaic.net who do a whole bunch of really cool things with rhino. The one I'm using creates an oscillating lattice around a curve. Perfect for the bridge.

Among all the options available is a min and max variation for the radius so that I ensure that people can still travel inside it. The results are below.





Hurray array... not quite

My original idea was to use the "arraycrv" command to place the cylinder module along the path at 3m increments (the modules are 3m after all). It should have been quick, easy and painless.

4 crashes and computer restarts later, I'm having another episode of rage against the machine.

clearly it requires too much cpu power. I checked the file size, and I guess unsurprisingly, it was 378Mb. The last two screenshots show that it worked with the structure for the cylinder, probably because it's a simple shape. No such luck with the cylinder skin.





Problems ahoy

So I had some processing problems...

it seems that creating these modules (especially the booleandifference of all the cylinders) takes up a darn lot of memory so I had to just do the first layer of cylinders, cut that section, array down 8 times then rotate as necessary. Even then I didn't have enough computational power to union them afterwards.

Rage against the machine.

Rotating the cylinders

I also want the holes to rotate so that they eventually align in their 3m modules. To do this requires some basic maths.... ohboyohboy...

there's 8 layers of cylinders to be boolean functioned.

360 degrees that they cover.

360/8 = 45

each layer of cylinders has 36 cylinders.

45/36 = 1.25

so each consecutive layer needs to be rotated at 1.25 degree increments (1.25, 2.5, 3.75 etc etc) so that when it starts at the next 3m of cylinder, it will line up with the front of the first.

phew.

Rhino modelling

As the post suggests, I'm bringing it into rhino where I've extruded shapes for the cylinder, the structure and later, the path. To create holes in the cylinder, I extruded then arrayed a set of cylinders (36 of them, so they're arrayed out at 10 degrees each) and will booleandifference them from the cylinder.















Final model beginnings

Making the sections in AutoCAD which I will later align with the correct position (the blue original rail is at the right position) using rotate3D and align.

The skin

The material choice is obviously important because of how it interacts with its occupants. I'm going for wood, which can be connected to the circular structure in panels (curved ones). They could also be pre-laser cut in modules like the work shown below.

The first are examples by David Dessens (l) and Daneil Widrig (r). Checkitout at http://www.generatorx.no/20080229/generatorx-20-update-2/



This one is by Tim Shork, you can find more about it at http://www.generatorx.no/20080128/generatorx-20-public-presentations-2/

How it works

In choosing the cylinder, it's left me with some questions about its construction. This is how I would tentatively propose it would work in section and elevation.

Even Sherlock Holmes approves.





Last minute uploads

just bear with me, will be uploading the rest of the process and the final renders now. It's to make it coherent for the A5 booklet as I'll be cutting and pasting it from the blog into indesign.

Oct 4, 2009

Cylinder variation

And here's the cylinder version. It suits the curved path more than the rectangle, so I'm going to stick with this one.



Rectangle variation

and here it is. A lot more convincing as a sculptural element.



Height within the bridge

As far as the rectangle and the cylinder are concerned, the rectangle is the easiest to calculate measurements. All I really need to do is allow for about a 2.5m clearance space and the 3m for the width of the path and that's taken care of.

The cylinder is a little more difficult, and relies more on approximation (in a sense). I've allowed for a 4m wide cylinder that should fit the path in line with that of the rectangle in the diagram below. If the path is 3m wide, it allows a headspace of a little over 3m, but this will shrink depending on structure.

Preliminary render

There it is.

A little unconvincing as a sculpture for now. Let's test it out with the rectangle and cylinder.

Capping the surface

Use the cap command in rhino to close the surface, and tada...!

Can we fix it? Yes we can!

This is me.

Correcting my stupidity.


A little mistake

Forgot that sweeping along a non-curved line makes monstrous, jagged results... so I had to trace along the original path's control points with the control point curve line. This is the result. Looks pretty good.

Variations

They're baaaaaack...

and they're happier than ever.

I wanted to test out some variations in how the bridge path relates to its skin. The first is an average bridge, with path and railing. The second is enclosed in a rectangular skin, the third in a cylinder.

Working in AutoCAD

I saved it as a .dwg file so that I could work on it in AutoCAD, this is to define the path width and depth (3m wide x .35m deep). Using the rotate3D command, and the align command, I'm going to align the middle of the path shape with the middle of the start of the ramp shape.

I'll import it into rhino to sweep/loft along the shape.

Smooooooth...

Just an update on the previous post, I rebuilt the line using 100, 200, 300 then finally 500 points. Between each change I used the "smooth" command to smooth out the line. Pretty happy with the final shape now.

Final path shape

Pretty happy with the overall shape, it takes into account the start of the location, which is just in front of the WA museum, and the curve over the road is to account for the implied force of the traffic, as I've covered earlier. May need to smooth out the curve though, but overall it's alright. Also, I worked in the top view to ensure that the 1:14 slope is maintained.

Rebuild it and they will come

Lies!!

Obviously the current path is inadequate for the sculptural qualities I'm looking for so I need to use the "rebuild" command in rhino to make some more control points. I'll test it with 50 for now and see how that goes.

Manipulating those control points...

Drag the points up specify the distance required (4.4m) and the bottom right panel shows how the points are being moved in the Z axis, while the top view on the top left panel shows that the path hasn't deviated.

Woot.

Control points

Turned on the control points so that I can manipulate the heights of the points. I do want the bridge path to be 4.4m above ground level obviously.

Working on the path

Ok... so... turning off the other layers makes it easier to work on the work path itself.

Surprise, surprise.