Equality problem

Hello everyone, it's been long time i've not used grasshopper and I tried to make a very nice tutorial from ThinkParametric for planerizing hexa panels with kangaroo.

I have a very strange problem, my equality is not working normaly, i don't undertsand why i d'ont have two consecutiv true ? If anyone has an idea ?

Attachments:

PlanerizeHexa.3dm, 180 KB
PlanerizeHexa.gh, 12 KB

Replies to This Discussion

Permalink Reply by peter fotiadis on September 29, 2015 at 12:48pm

Planarity is somehow a chimera, he he. If K goes a long way ... chances are that you are after your tail for good.

K2 requires w7 and up (since the .NET required is not supported in older versions). Can co-exist with K1. Works in 32/64 bit O/S, Rhino SR9 and up (but that is tested using Xeons that are 64 bit anyway).

Here's a rather strange case (K2).

Attachments:

Planarity_on_Polygons_V1.gh, 26 KB

Permalink Reply by peter fotiadis on September 30, 2015 at 12:23am

BTW: Leaving aside planarity ... and focusing to that "equalizing within boundary" problem ... this is in fact kinda a tensile membrane "equilibrium" situation:

1. "star" lines try to handle polygon's "boundary topology".

2. springs try to pull/push the whole combo around.

3. polygons are pulled against a guide surface (as a mesh actually because is way faster).

4. properly defined perimeter points pull against the outer/inner Loops (BUT they are NOT anchors). If you sample ALL the perimeter points you are shooting your feet.

A fine balance: if you get it wrong > bad things happen.

Permalink Reply by Nenov Ivo on September 30, 2015 at 5:06am

Second Code is even nicer !! I will try experimenting some things while combining it with inflated or tensile structures !!

Thanks for the clear explanations.

So you say planarity is not possible or it's very complex ?

Maybe we could combine kangaroo with a genetic algorithm that modify the shape and the parameters of the different forces to reach a maximum relative planarity. With K1 seems hard cause modifying the force parameters doesn't affect the solver when it's already on or not in a good way... maybe with K2 ?

I've just installed it and it's a whole new world for me (to be honest i don't understand anything for now hehe)...

For the processor things i have laptop with intel Quad Core i7 - 4700HQ. 2.4Ghz but it's a shitty Asus comp for gamers so it's very laggy, blue screen problems cause of the two graphic cards that are fighting against one another... When I use too much kangaroo Rhino crashes...

Also, I'm wondering if it's possible to inflate a hexa grid and relax it at the same time ? because i managed to make it (with your code) but with two K solvers. First i inflate the planar grid with a solver and after I relax it with another one.

BTW i love my tail !

Attachments:

InflatedStrucutureRelaxed.gh, 46 KB

Permalink Reply by peter fotiadis on September 30, 2015 at 6:45am

With regard some proper workstation: If at some future time you have ~ 2K to spend (or a friend from US/Canada) drop a word.

With regard planarity1: it's VERY expensive (real-life, real-projects that DON'T leak for ... er ... a certain period of time).

With regard planarity2: See def attached: Not related with hexagons et all ... but useful in order to puzzle yourself about what planarity could be in real-life. Made for a fellow user recently who has YAS type of stuff in mind.

With regard your hexagons: What exactly is the system that you have in mind for doing it in real-life? (LBS description, Aluminum series [in depth], Panels [glass, plexi, mylar, polyC etc etc]).

Plan Z : Skype.

Attachments:

YAZ_kinda_demo_V2A.gh, 141 KB

Permalink Reply by peter fotiadis on September 30, 2015 at 6:51am

And the rhino file

Attachments:

YAZ_kinda_demo_V1.3dm, 183 KB

Permalink Reply by Nenov Ivo on September 30, 2015 at 8:33am

For now I don't really know what I want... I have some "goals" too but not so "sainte-ni-touche", prefer some more ghetto gyals hehe ;).

I want to make something with kangaroo this time. I don't know what for now. Last year i discovered the power of genetic algorithm with galapagos so i made some mid-wood structures. This time would like to experiment something tensile or inflated... Really like Achim Menges and Andrew Kudless structures.

I don't want to copy them(and i can't hehe )... want to find my own thing.

And for the material will be wood or cardboard cause at school we have a laser cutting machine.

It's just for fun not for money, not for glory !!

I see with your example that having planar hexa it's difficult but i think dividing them in 6 triangle is the solution.

And thanks again for the codes and the propsitions ;) !!

2k in maybe... beeing optimistic... 2 years...

Permalink Reply by peter fotiadis on September 30, 2015 at 8:58am

Well ... the danger here is to get used in this Academic top-bottom way of thinking and have a very bad landing (actually a crash) when at later time ... well you know what I mean > hello cruel real [brave, new] world > ... > pain + tears + why this is happening to me etc etc.

Thus the only advise worth considering is: try to see things "in parallel" with the other way: bottom-top that is (what real nuts and real bolts are actually required in order to do this ... and how many dollars they cost ... blah, blah).

This is not the easy path, but is the only path.

BTW: One of these days I'll mail you a "weird" paper of mine: the Aurora Project.

Permalink Reply by Nenov Ivo on September 30, 2015 at 7:55pm

Thank you very much Peter for your help.

I tried to copy Matsys ShellStar pavillon and I'm very happy with the result ! But I have to optimize the planarity of the cells more... tomorrow. Any advises are welcomed.

Attachments:

MC_3.gh, 157 KB

Permalink Reply by peter fotiadis on October 1, 2015 at 12:03am

Hmm ... since the top is resolved ... let's get used upon what bottom means (pain + tears + why this ... blah, blah). I mean some bottom that protects you from the elements, that is.

Here's the most efficient (i.e. doesn't leak) transparent "roofing" aluminum system known to man (designed by Greeks) - it's a 3dPDF thingy thus use the latest Adobe Reader. Study the whole approach: why these "hinges" are around? What type of issues addresses this freaky combo? Who other than a lunatic could design such a "simple" system? Why this is happening to me?

Used the structural glazing version which is the most expensive and has the shortest service life (but life's short anyway > spend zillions > get the best, he he).

Moral: pile of worms.

Attachments:

ALUSTET-MASTER-45R_not_restricted_000.pdf, 151 KB

Permalink Reply by Nenov Ivo on October 2, 2015 at 4:07am

The detail looks really nice (and clever, maybe too much...).

If we immagine the kind of structure I've post above, we need a metal structure describing the form and supporting the double T elements that are orienting the metal carpentries ("menuiserie métallique" sorry for my lack of english) with the hinges which are supporting the glass surfaces...

Things I don't understand for a structure like this:

1.How the elements are assembled on the building site ?

2.Do we have to consider that for each hexagonal glass we have only two curvy links that are facing one another ? (if more it connot be assembled !)

3.For the almost flat hexas we make hexa glasses, for the others we have to divide them in 6 triangles... ? triangles glass is very expensive but who cares hehe ...

I'm asking too much but it's just for my personal curiosity, I don't think i'll use this system because it will just make some nice 3ds for my portfolio but I want MODELS (prefer real things not virtual).

For the materiality, I was thinkings about wood surfaces assembled with a mid-wood system... It would be possible if I use a soft wood that accepts deformations...

I'll post the WIP when I'll simplify the main structure. Now the thing is composed of 7 triangles... I'll first start with 1 and complexify after hehe.

Permalink Reply by peter fotiadis on October 2, 2015 at 5:01am

Hmm ...

1. Buy some suit (Armani I do hope) suitable for desert climate (portable A/C is a must) and catch a plane to ... then I can teach you "on-site" how to do it, he he.

2. This variant is not adjustable (i.e. for LBS construction tolerances THAT ARE THE NORM). Thus it is used for triangle panels (the right thing to do ... or see def attached). The adjustable thingy is ultra complex (and expensive).

3. Tempered hex glass panels (triple layered + the thermal gap) == GNP of Nigeria. Pay zillions get nothing but Vanity.

At this delicate point I would suggest a break from all that stuff.

Get this def (Load R file first) and study the way to get flat panels from hex thingies. Actually flat "sub panels".

This is not as stupid as you may think: from some dimensions and up ... talking about hex glass panels is out of question. If on the other hand we can achieve planarity with trapezoids derived from the hex thingies ... why we should bother with hex planarity? And since there's no panel equal to other ... why we should bother with "relaxation"?

Thus ... consider this as the plan Z(ZTop) for that issue of yours.The bad news are that there's no components around BUT you can pretend that these came from some sort of GH plug-in, he he

Attachments: