For distraction you no longer need to switch which window you're working in

A MITCH (ha) needed update

Last time, on Mitch, we saw various plants which I intended on 'biomimicking,' including Mangroves, Banyan trees, Whitebark Pine, and my personal favourite, creeper tendrils.

As I explored my options more, I increasingly seemed to be turning to the creeper tendrils, which seem like a great opportunity to emulate for my bridge. The twisting of the tendrils allow for a much greater rigidity than the minimal form of them would suggest.
As I moved forward with this design, it began to develop as a bridge which literally twisted in the same manner as the tendril, but as a bunch of them.

The central sketch in this image shows the thinking behind this. It would be a series of rings, with steel rod supports between to provide the overall structure. These supports would be connected with details like the ones shown below.

I had even begun to develop a script that would allow the bridge to adapt to different conditions, which may be an unnecessary use of grasshopper.

Upon further though, I felt that the number of tendrils was beginning to take away from the concept, and it was beginning to look unlike a creeper at all. I took a step back and began to think about how the plant actually supported itself, leading me to two solutions, twisting, and the cross section of the tendril itself.
This, however, was abandoned. Going back to twisting, I began to look at tensegrity structures as suggested in a critique.
Like this, but on its side. I like that tensegrity structures allow for a great deal of strength with relatively little structure, much in the manner of the creeper tendril. So, this idea is being developed, as can be previewed in the following details.

Progress Report 2

Alright, so, being silly I would always check the blog but never post. So here’s to make up for it. Be prepared, it’s long.

Having pursued the “snake” as a model to emulate i was unsure of what to take away from it. I found interesting videos on snake movement and side winding and the mechanisms employed for locomotion. This turned out not to be solely a directional movement in one plane, but also a vertical displacement of their body at two points.

There was no foreseeable application of this mechanism unfortunately. What i did admire and find useful was this elegantly and contortive ability. This led me to think that if there was a program i hoped to address, how could one designed element be repeated and nested in a manner that would satisfied the geometry of the site in an elegant manner. This is similar to the idea of the wooden snake.

I guess that reminded me of Grimshaw’s train station and its unique winding and widening condition that was responded to parametrically so that one truss was designed and could be locally configured in each case along the station’s length.

At the same time i remembered some explorations i found some schools doing that resonated with my interest in the snake and nesting.

Some other biological models i was looking at included:

Reviewing some of my explorations with Vince he reminded me that the snake investigation was more akin to the armadillo. Depending on the number of bands that it has the armadillo has the ability to curl into a ball.

The nesting properties were interesting in that band width (is that where “bandwidth” comes from?) varies, and these bands are connected between each other with skin, as a membrane. This type of “scaling” effect also reminded me of fish scales, artichokes, pine cones.

This immediately reminded me of Achim Menges’ project Responsive Surface Structure which examined how the material properties of pine cones allowed for humidity to deform scales. The veneer strips in this project also curl up due to their material properties.

Remaining with the canopy idea, i was inspired to see if nested components could possess ever changing degrees of nesting as a product of a membrane that would link them and respond to climatic conditions. Maybe when it’s colder the membrane would shrink and bring the nested components closer together to delay the escape of heat from the space beneath. Maybe it would be a question of how the steel components would expand and contract.

At this point i started taking the context conditions more seriously to see how this could affect my intent. I discerned that most platforms for train stations are rather straight and so i wouldn’t be able to find the twisty condition i had hoped for as with the waterloo station. So i located a site in Vancouver that was really interesting contextually. The Skytrain is the raised LRT system in Vancouver and the Main- Science World station is a unique site: an interesting neighbourhood, raised LRT platform, intersects a building, is by the water, near a geodesic dome, and has a canopy that has been acknowledged by the city as “needing a facelift”, providing a good opportunity to present an alternative.

Busyby’s Bentwood station in Vancouver was a source of inspiration and information for hwothese stations function.

This all led to a clearer intent with my project:

Who

The skytrain passengers awaiting their train will use the structure. The neighbourhood community will identify with this transportation node.

What

A steel skytrain shelter that is inspired by the properties of the armadillo scute or set of banded armoured shells. The shell that is created is manifested in such a way to mediate and manipulate the environmental conditions of which the shelter is situated in, adn accommodate a public platform that is oriented towards the shorefront.

When

The structure through its materials and construction and architectural presence should be long lasting and become a monument that will survive even if the skytrain does not.

Where

Main-Science World station in Vancouver, BC, is an existing elevated Skytrain platform that hosts two skytrain lines concurrently which intersect the VanCity building. The canopy and platform structure will be reworked to also take advantage of its proximity to the shore and the science world geodesic structure.

Why

The unique context within which this station is located, and its relevance as a station for a community in a unique neighbourhood, demands an equally as vibrant architectural identity. The nesting and configuration of the armadillo shell is relevant to the construction of a shelter and may generate a means for a unique intervention.

How

Through the use of structural steel and other materials the elements of the structure will be arranged as a response to sunlighting conditions and to hopefully accommodate thermal containment or venting during seasonal requirements.

I felt a more scientific and process could yield a means of determining how to shade the space using grasshopper definitions at the following source http://www.tedngai.net/

I located resources that could help me link environmental data for specific sites to geometry to be paneled onto a form of my own making. First i was looking at one that used ecotect as part of theworkflow, but i don’t know how to use Ecotect so that didn’t work out that wellf or me

Found one that accomplishes a similar task in Grasshopper.

http://www.tedngai.net/experiments/incident-solar-analemma.html

“This definition displays the accumulation of direct solar radiation throughout the year in false color.” It does so on a surface and so i then linked this rgb mesh output to a definition that tiles 4 different types of modules based on what i‘m going to assume are gates between certain rgb values to determine which module to use on that area of the original mesh surface.

I did a test with some less relavent geometry to this project, but the geometries were i guess too heavy and i spent lots of time waiting for grasshopper and then it would end up just freezing, hence why it looks incomplete. But i assume i‘ll require fewer complex geometry and at a larger scale for what i want to achieve. Using box morph in this definition though doesn’t really allow for the nesting i seek. Maybe there’s an easier way to do it? Started looking into paneling tools today... i wonder if theres a way i can link may data to it? Maybe i’m over complicating things? I’ll have to investigate a bit more i guess!

And as a measure of our productiveness in studio i have photos:

Desk Update:

Grasshopper Hour - First results

So a few of us have spontaneously banded together for one hour bouts over the last few days to brush up on our Grasshopper skills and challenge ourselves to create some interesting and responsive definitions.

For those unfamiliar with ghoppah it is a free plugin for Rhino that can be used to parametrically control geometry for rhino. this means creating relationships and systems for how this geometry is generated and/or manipulated based on certain rules. I think that's a decent way of explaining it at least.

So our objective with the first exploration was something that Perry was trying to explore for his groups Ontario Place Collaborative Exercise submission. The intent was to be able to control a grid of squares so that they could gradientially transform relative to a point in both their shape (from square to circle) and in size (from large to small).

Some of you may be familiar with the now cliche use of attractor points.

This was the basis for controlling the definition. You can easily find this definition on the internet.

Instead of circles we figured we could take the edges of square and remap the distance values between the attractor point and each grid cell point to retrieve values to control the fillet. The relationship between the edge length and the fillet was x/2. This worked successfully, though in retrospect i think we were all expecting the filleting to appear more dramatic as an effect.

At first we thought we would be scaling all the squares relative to the point AFTER the filleting occurred, but realized the whole grid would also be scaled. Then we figured what if scaling took place first (again, a function of the remapped distances to the grid cells) and then the filleting took place. When we rearranged the placement of the square component in the definition, we also realized there was a fillet option in there. So we connected all that stuff up and unfortunately the reverse operation was taking place: we had hoped the geometry would scale down relative to the attractor point, but instead the smallest geometry was exactly where the attractor point was. Jordan then showed us that reversing the list for the fillet input worked to achieve what we sought. What still doesn't seem to be working is that the geometry that is closest to the attractor point doesn't generate a square with sharp corners. Not sure how to resolve this just yet.

Ask me for a copy of the ghx file if you want to play around with it. Can't seem to find how to

post up files here.