March 2009 Archives

Going back to the TS renders to put together the plates for tomorrow's tutorial. The idea is to make these a bit less we're-standing-in-the-middle-of-the-desert-y and more New York-y.

Good tutorial, as always, with Wolf today, though it was pushed back by more than two hours which messed up my day a bit.

Have now got a better understanding of how the different pieces of research that I've been working on go together, from thrust line analysis through to funicular model. And an idea for how that actually feeds back into the corner.

Need to crack on. A few worked-up versions of previous pages + other work in progress. More to follow after what I suspect will be a long night. Will work on the big model at the AA from 10am tomorrow.

What it looks like before hanging the chains. Good meeting with Mike today, and good comment from John: 'start loading the model from the bottom up, that way it will be easier to understand how the design of the lowest arches (the ones taking the most loads) needs to be rectified in order to deal with whatever asymmetries are above.'

Scale 1:100 (taller than I am, though admittedly that doesn't tell us very much).

Working out the weight of the structure and how the arches should be connected in the model, looking at material properties, and breaking the tectonic system down into eight box-like segments that are easier to model and calculate. Work in progress, etc, etc.

One of the theory pages I'm working on at the moment - more to come later on tonight. This one shows the difference between using catenary and parabolic arches in terms of stress, as well as the difference between uniform and non-uniform sections.

Working on digital + physical model and not always at the computer, will take a few more hours before I blog again.

 

More screenshots from modeling the arches: changing the initial geometry to what I thought were roughly catenary curves, then turning those curves into actual catenaries (and realising the initial attempt was rather wrong), thinking about how they may go together in better ways, then remodeling them, then coming up with new tweaks, remodeling again, etc, etc.

Possibly overstructuring it, but I suppose that's better than the other way around at the moment.

The next step tonight is to turn this iterative process into a legible page (sorting and annotating these screenshots; I think they will be published so that they should be read from the bottom up on the page below), then sketching out the whole presentation more precisely, then starting to draw the first section from this study. Simultaneously looking at catenary systems at large + different strategies for how to make curved structures using brick.

Need to push on. More to follow tomorrow.

Working out catenary arch shapes with 3D Studio Max's Reactor (physics simulation tool) Rope command, which, given certain parameters, produces catenary curves.

Diagramming up the structural system, then beginning to turn that into the final 3D model. Next step is to 'brickerize' what I have here, but haven't come that far yet.

To be continued, obviously.

First half of the new A2 - more to follow soon.

Seem to have caught the Dip 9 bug again and have been down and out for most of the day sweating and shivering under blankets, but am slowly coming back to life. Here's a first sketch of the principles that I've discussed with Wolf + a page showing the progression of corner moments from the first (massing) model, through the generically sinuous articulation through to the final iteration. More to come during the night.

The self-imposed deadline of... now almost worked out, but not quite. I'm still updating the A2 fold-out, while trying to work out structural ideas and finishing off the generic and specific versions, but this very minute have to leave the computer. The three pages below show how I started (with a quick analysis of the rules governing Hugh Ferriss's renders of the 1916 zoning law envelope) and proceeded to use those same rules to create the new orthogonal envelope, which I'm now modulating with my n-point corners - and finally how the fold-out is beginning to come together (editing out superflous info, etc). Sorry that this is a bit rushed; have to run. It actually looks like less than it is - the drawings in process not quite bloggable but basically a matter of connecting dots. Just takes time. I'll continue sketching from here and will work through the night tomorrow and have a lot more come Wednesday.

Almost there with the first A2 half of the fold-out drawing:



Hope this is beginning to give a clearer picture of what I'm trying to do. Progress is a little slower than I had hoped, as it's tricky to find the right balance between the two previous versions (the repetitive one vs the much wilder one), but it should be in place for tomorrow. The second half is quicker as well. Just remodeling the new envelope as we speak, then need to add a structural 'x-ray', and pull out the primary corners I'm looking at. Along the way, I'm making drawings such as this one, to explain the steps (this shows how the corner can be modulated through bifurcations into 2-6 points with respect to structure, openings, and enclosure):



I probably have to remake this a little later on as some of the examples are a bit wonky, but I reckon it's good enough for now.

Next is just a quick sketch (nb: not even a drawing yet) of how the 'sinews' might grow into column-like structures. These are too thin, but give an idea. It quickly becomes very neo-gothic when the corners are concave, which can be interesting in places, but I like the previous convex version better, so will try to rework this into a version that is primarily convex, with some concave moments.



Now just workingworkingworking to get things done for Wolf. I'll try to blog more final versions later on, though it takes a bit of time to do this, and now I just want to push on and get this running faster than it knew it could run.
 

Quick update: sketched out the pages I need for my tutorial with Wolf tomorrow. Remodeled the existing massing model again + adjoining buildings to get a proper hidden line axo. Started getting the first drawing in place. It's meant to be a fold-out (2 x landscape A2; we'll see if this actually works or not) content page showing the development from section of existing building through to final new (exterior) form, indicating where my new (primary) corners sit and how they have been informed by the precedent studies. Need to push on, will blog more later.

Working towards the first version of the façade/structural treatment. This is perhaps a bit wild - the next step, what I'm doing now, is to place three carefully chosen corners inside the bounding box and then connect them in this same way. These shapes are based on the notion of dissolving (or negating) the corner; the curves shy away from any sharp moments. Instead, all orthogonal corner action will happen inside the building.

Super quick renders - for some reason, my computer hasn't got enough strength to turn these into linedrawings (any ideas for how that can be done? Make2d just crashes it). Tutorial with Tony at 3pm, then back home to crack on with this and get it done for tomorrow.

* The problem with sinews being that they are strong in tension and straight, whereas I'm working with curved elements that are weak in tension - though since I haven't got a better term at the moment, I'll just stick with it; we know what we're talking about.

"The only thing we have to fear is fear itself - nameless, unreasoning, unjustified terror which paralyzes needed efforts to convert retreat into advance." /Franklin D. Roosevelt

Yes Franklin. Say hi to the past four weeks.

Trying, then, to sketch and sketch and sketch to get past this threshold and move on to more interesting things. My setback corners smoothen out the pointed exterior corners so that all visible corners on the outside are dissolved - the building hides the corners on the inside. Now working on translating these hand sketches into axos where the connecting curves thicken into what we have called 'sinews'. Not there yet but thought I'd post a quick update; will post again later on today. Straight from the desk:








And here's the first step of turning these into actual 'sinews':



As I said, better version of this latter one coming up later on today/tonight. Comments, suggestions, friendly pieces of advice much welcome.