Help bending an object (not surface)

Replies to This Discussion

Permalink Reply by peter fotiadis on November 24, 2015 at 8:38am

Hmm ... I have something rather "similar" but it's made solely with code (I guess that you don't want such a solution, he he).

I'll try to fix your def later.

In the mean time:

1. Is this academic?

2. "Ideal" fins must have some more "rounded" profile right? Then after placing them properly in space (Orient a given profile across planes "along" guide curves) you want to apply sweep1 on these? So ... it's NOT bending ... it's just sweeping using the proper paths (curves).

3. What are you thinking with regard making the "fins" in real-life ?(even "making" - in theory if 1 is true).

4. Why the K2? These things (sweep breps) should "follow" curves derived from a given envelope that "engulfs" a given building (according some rules - for sure - that have nothing to do with a physics engine). Or you attempt using K2 for "bending"? (utterly pointless).

Permalink Reply by Katelynn Tate on November 24, 2015 at 8:59am

I do not know the code stuff at all yet, willing to learn but that seems to be a whole separate monster! so I may be best off learning all of this first,

It is not academic, trying to learn while work is slow (is this a academic advise only board?)

1&2 - The fins- you probably saw the sketch with the profile, if not take a look at that. I want the more square part that is to one side to be the structural component, probably steel and it would have to be more rigid, likely smaller pieces welded together to become a sort of curve, I do know that but didn't want to have to dive into super detail. The part that then is the fin could be a thinner metal, that would be like a skin over that column and if thin enough it would be bendable. The idea is that these fins are on the exterior of a building, like an overall skin but are completely separate from the buildings itself. They allow light in depending on what direction you are looking from and the light is never direct. The profile likely would have more rounded edges but I don't think it is completely necessary, this isn't something I played with too much yet.  I think this answers questions 1 and 3 and gives you a little better idea of where my minds at.

From what you said in question 1 are you suggesting I put the fin profile everyplace I want it to start and end all manually in rhino and then sweep from one to the other? This could work I think, and would pretty much be all I need to get the forms I'm going for. One of the issues I am having with that is how to loft on a curve. I keep getting and end product that just is not behaving correctly or that turns itself into a straight member instead of curved to get from one profile to the other.  

3- If by K2 you mean kangaroo 2 that is because I downloaded it not knowing what I was downloading :) I keep seeing these kangaroo definitions that are similar to what I am wanting so I just downloaded the latest and didn't realize there were two different plugs basically. I had just downloaded K1 before answering you so hopefully I will be a little more in the loop now. If you don't mean kangaroo.... then you'll have to tell me what you mean. 

Permalink Reply by peter fotiadis on November 24, 2015 at 9:57am

Well ...

1. This Forum is certainly about helping people but helping pros or future pros or providing solutions with regard real-life cases ... I do hope that you understand what I mean.

2. With regard placing the fins see def attached. The "only" thing MIA is: (a) elaborated ways to design the profile, (b) a policy for creating "parts" (for fabrication) out of these rather big mono block fins, (c) an optional policy for "twisting" the planes used for orienting the profiles and ... er ... (d) the big thing: the "adaptive" surface that yields the sweep1 rail curves (see attached).

3. With regard doing an "adaptive" (so to speak) surface that "engulfs" a given building according some rules (and taking into account ways to "split" the big fins into "parts" + the required supporting system) ... this is certainly doable but if I'm going to give you some hints it would be solely via code: i.e. (a) creating a "gross" engulfing surface (or many), (b) modifying the control points with a variety of ways: proximity/altitude or other, (c) "testing" the "parts" for validity etc etc.

PS: Load Rhino file first.

Attachments:

• Fins_for_skin_V1.3dm, 179 KB
• Fins_for_skin_V1.gh, 19 KB

Permalink Reply by Katelynn Tate on November 24, 2015 at 10:41am

1- Yes this makes sense I just wanted to be sure it wasn't student only as I had not seen anything that said that, but you never know... I would love to be considered future pro haha but I am just not there and this will not be built. This is a personal project to further my own learning in the software and how it can be used to become more believable architecture.

I want to be sure I understand what you did here- you made a surface that defined the path and then used prep frames to break it into even segments that will be the spacing for fins and then projected lines onto the surface where the fins would be place. Then you used prep frames again to divide the curves into a number of segments that looks like it will help determine the orientation of the fin profile. Is the more frames here a closer fit to the beginning surface?

Then you developed the curve for the fin profile, rotated and scaled it how you wanted then set them to the orientation defined by the previous prepframes. Then you used a sweep to go through the fin profiles along the curves, using them as rails.

One of the things I do not understand is the modifier buttons that are sometimes applied, like simplify and graft. I will look for a tutorial that better explains these and why it must be done sometimes (in this instance the sweep does not work without the graft) Also, are the prep frames used quite often? I had not yet seen them in anything I watched but it seems much more logical than the way I was trying to go about it.

When you start to talk about rules, adaptive objects and such is this mostly all done via coding? It seems as though all of the buttons and such in GH similar to the ones you used for the definition are all things that could be done in Rhino and the only reason to use GH is that it makes it easy to adjust things.

Thank you very much for your help, I think I can get to my desired end point from this. I think I will start to watch some tutorials on the basics of coding though as much of what I want to do is within that I think.

Permalink Reply by peter fotiadis on November 24, 2015 at 11:15am

Well this case is a giant pile of worms IF real-life stuff is to be made. Is way far more complex than it appears and requires some serious engineering (even if large scale 3d printing is used; stuff that greatly facilitates "impossible" tasks of the past).

Anyway..

1. Graft means get a List and make a Tree thus the sweep has profiles (as items in tree branches, i.e. Lists)  VS rails as ... er ... we need to match apples to apples here: for each branch that has profiles get a branch with a single item (the rail).

2. In general without fully understanding the way that GH manages "nested" collections (DataTrees) is utterly frustrating to attempt anything. Spend some time on that matter. Some fellas would advise you to learn "on the way". I would strongly advise the opposite: first understand what a piston is and then design a machine.

3. Rules et all ... well you can do some stuff via components but the things that I have in mind require code not because is far more "efficient" in complex data structures (a matter of personal preference that one) but mostly for controlling items in collections on a per item basis (don't ask what this means, he he) tracking history, using hierarchies of instance definitions (blocks), controlling sliders, controlling events ... and many other ultra freaky things. 

4. Coding is the next step towards the Holly Grail. Requires serious commitment, time and guarantees frustration, agony, and pain until ... you arrive into a point where all things appear "easy" (kinda like attempting a double forward in windsurfing: easy provided that there's "some" scars around). 

Permalink Reply by Katelynn Tate on November 24, 2015 at 11:23am

This makes sense. I think graft is something I have been needing in several cases so very glad that was cleared up!

I am learning the nesting stuff can be very frustrating and agree that it is something I need to spend time learning. It seems as though many skip over explaining what all goes into the nested parts and many newbies are stuck trying to figure out why they can not get similar results. Luckily I have access to Lynda tutorials and they seem to be pretty good at explaining things thoroughly.

I will start looking at the coding stuff. As terrible of a headache that it sounds like it is you can achieve some really awesome stuff.

Thank you again for your help!

Permalink Reply by peter fotiadis on November 25, 2015 at 12:01am

Later today I'm going to add some stuff in this "guide" def provided (Using straight Loft: thus making straight "segments" that make much more sense cost wise).

In the mean time:

1. Parametric combined with coding gives you enormous power but that is a malediction: power brings corruption and absolute power brings absolute corruption (thus be extremely skeptic with regard what "awesome" could mean - in the long term).

2. For instance: there's no reason to spend lot's of dollars to make a WOW shading "envelope" (in your case works only if the axis is North-South and the surfaces point East-West) that would never reach an investment break even point. 

3. For instance: remember these? Humble stuff used in the good old days (short brackets "protrude" out of a semi-structural curtain wall and support aluminum "fins" [horizontal-vertical] - in this occasion shown rotated via step motors):

4. And why such solutions are considered obsolete/outdated/"old"/why bother these days? Because absolute power brings absolute (mind) corruption, that is.

5. And if you add this to that ... we arrive into YAS Hotel type of "WOW" situations: get an ugly thing [the hotel itself] and spend zillions to make it "better" (worst in fact).

best, Peter

 

Permalink Reply by Katelynn Tate on November 25, 2015 at 10:45am

Many of the parametric and coding objects/buildings that people have come up with are nothing but really interesting pieces that can never be realized. It is a nice thought though and maybe one day this super elastic/see through/structural/pliable material everyone makes these things out of will be real. They will continue to be used in design completions and such that have no intentions on being built.

I think things such as your basic fin shading devices are fading out for the moment because everyone is looking for the next big thing and wanting to stand out. Many buildings have been successful finding it but it costs a ridiculous amount and has to be completely custom fabricated so in most cases it is unrealistic. So many companies want to build that iconic building so that everyone recognizes their company that something so generic/common just is not what they are looking for. The YAS Hotel is a prime example of this.... it's not a good looking building but everyone will know what it is.

Permalink Reply by peter fotiadis on November 25, 2015 at 12:46am

BTW: Spend some time on that one attached made long long time ago.

It makes ...er ... hmm ... "abstract gills" using surfaces. Has some historical value as well: one of the very last defs that I made using native GH components: now all these are "absorbed" into a C# that does 50 times more things - but you are not ready [yet] for that type of freaky stuff, he he.

BTW: if these things are made out of, say, poly carbonate semi transparent sheets [that brinks us serious packing issues in order to laser cut the pieces with the min waste material] and IF you can design a proper support system [like an ultra cool and expensive planar glazing thingy: Google that]... then ... well ... that is a far more "economical" way to make some "shading" envelope (that is NOT required anyway: Google "smart glass" and the likes).

BTW: "thick" poly carbonate (~ 10-15 mm) can support itself quite well not to mention that it can "bend" (but looks a bit kitsch over time - ultraviolet effect etc etc) ... the other option is to spend millions and make all these with sandblasted tinted glass.

Moral: absolute power ... blah, blah

Attachments:

• gills5.gh, 172 KB

Permalink Reply by Katelynn Tate on November 25, 2015 at 10:54am

I have no real idea how these things would be made.... which goes against what I said about wanting to know how to begin to use grasshopper as a tool to make "real" architecture....maybe I will figure that part out but for now I am still just having fun and playing with the form, the hard part can come later.

These gills are very cool, I will have to dive into it later this evening though. It reminds me of the Aros rainbow panorama just by being rainbow and rounded.

Permalink Reply by peter fotiadis on November 25, 2015 at 3:37am

Here comes the pain: the V2 that is a "bit" more complex than the V1 (but the forthcoming V3 is a nightmare: the ideal thingy for a novice, he he).

PS: Added some future "tasks" for you in case that you have some Dark Side stuff on sight (but trust me: think it again).

PS: Obviously if the "fin guide curves" are divided into equal length segments - see option: Loft + straight Loft ... well ... that's half the bacon since we need some policy to yield "evenly" distributed "equal" fin components (that's what V3 does).

PS: Test surfaces are internalized (I do hope that GH can do that, he he).

Attachments:

• Fins_for_skin_V2.gh, 156 KB

Permalink Reply by Katelynn Tate on November 25, 2015 at 10:56am

This took the other one up like 100 levels.

First question out of the gate - You say to put my own surface in but "not with trimming info." What does this mean? like make a surface in Rhino but if I trim it in any sort of way it is not usable?

• ‹ Previous
• 1
• 2
• Next ›
• Page