Connect 2 surfaces with smooth transition

Hi,

I´m trying to connect two surfaces smoothly as illustrated in the attached figure. There is a component to connect two lines smothly, by using "interpolate (Intcrv)", I was wondering if there are something similar to that for surfaces or if anybody know how to do this.. Plz help !

Attachments:

Connect2surfaces.png, 84 KB
Connect2surfaces.gh, 23 KB

Replies to This Discussion

Permalink Reply by peter fotiadis on May 14, 2015 at 10:06am

Create suitable paths (the blue) rather using the tangent curve component (on edges and AT LEAST in one section) and then use sweep2 (using brep adjacent edges as rails).

Note: G2 continuity on paths is the min requirement for a "smooth" bridge.

Permalink Reply by Frede ranjit on May 14, 2015 at 11:02am

Is there a easy way to create a suitable paths ?? as a interpolation of the curvature of the surfaces?
I don´t understand the note??

Permalink Reply by peter fotiadis on May 14, 2015 at 11:16am

OK, I'll prepare a "simple" tutorial

Permalink Reply by Frede ranjit on May 14, 2015 at 11:18am

Ohh I see, i need a G3 curvature, but I´m not an expert in GH, so i don´t know how to establish this so it fits my surfaces..

Permalink Reply by peter fotiadis on May 14, 2015 at 12:36pm

Well ... G3/G2 are NOT always achievable (for obvious reasons). Anyway I'll completely reform your def using 90% less components (and some other tricks of mine). If I could use C# that would be far easier ... but I guess that you don't want such a freaky solution, he he.

In the mean time ... get an indication upon that tangent curve thingy (G1).

BTW: if this is some building ... G2/G3 are not really required (nobody could tell the difference). If on the other hand is a decorative object (with shiny final coating) of some sort ... then yes we should treat the "bridge" that way. Not to mention that in this case ... we don't need the bridge at all (it's all about how to design the whole "canopy" at-once, you know).

Attachments:

Connect2surfaces_NEEDS_TOTAL_REFORM.gh, 57 KB

Permalink Reply by Frede ranjit on May 14, 2015 at 1:42pm

I know, it was a messy script. Exactly, I'm not that advanced in C#..

The surface is used to create a grid shell structure at the end, so it is may be not that import whether it is G2 or G3.

Thank you sooooo much for the help and looking forward to the rewritten script :)!!

Permalink Reply by peter fotiadis on May 15, 2015 at 12:56am

Hmm ...

Before the V2 a couple of questions:

0. Is it some project that inspired you to do something similar? (post Images).

1. Are you trying to "morph" an "uneven" figure of 8 with a half "uneven" 8 acting as the height "guideline"? (section in Z so to speak). If so has any meaning using sliced portions of spheres and then attempting to bridge them? (I mean: is any reason for an exact arc [per shell] as a footprint?).

2. By grid shell you mean that you want to create a truss "following" the resulting brep? [polysurface if we bridge the 2 cells]. What type? (MERO: meaning struts/balls, FLAT: meaning a geodesic collection of struts/connectors or ABSTRACT: some kind of "liquid" thingy with/without W "thickness"?).

3. Is this a building roof/canopy/hangar/whatever of some sort?

4. Is this academic?

5. By "not that advanced in C#" you mean that you have some serious plans to learn the language (path to hell that one, he he) or you are just curious about?

6. If in the forthcoming V2 some stuff is performed via C# (meaning terra firma for you - at least at present time) could be that a critical issue?

7. Do you have GH TSplines 4 (Beta) installed? Are you familiar with that thing? (and the numerous bugs as well).

8. If Kangaroo could called on duty (Plan Z) for the upwards "morphing" job (starting from a "flat" mesh approximating an uneven figure of 8 ... etc etc) could be that an issue?

Permalink Reply by Frede ranjit on May 15, 2015 at 1:55am

0. It is a academic study, where a regular dome was studied first and now I wanna study a more free form shape. (The regular dome is attached+script).

1. Yes, something similar. I'm actually trying to "merge" two domes together with the negative Gaussian curvature at the bridge. The hieght of each dome is supposed to be variable, meaning different span to height ratios are used.

2. Yes, I want to project a flat grid onto the surface. To imitate a MERO system.

3. Is supposed to be a roof structure and could be used as a stand alone structure.

4. yes a student project.

5. haha, I'm curious about it, but don't have the time to learn it for this project.

6. I don't understand the C# language yet, so it would probably be an issue understanding the process.

7. No i have not?

8. Do you mean to perform a dynamic relaxation? and thereby get the desired shape?

Attachments:

dome.pdf, 228 KB
Regular dom, triangular grid.gh, 14 KB

Permalink Reply by peter fotiadis on May 15, 2015 at 3:07am

OK, required info gathered. I'll be back (in the mean time get an one way trip to hell with no return ticket).

best, Arnold S (Terminator 2)

Attachments:

dome_demo_v2.gh, 305 KB

Permalink Reply by Frede ranjit on May 15, 2015 at 3:04am

A quick question. I incorporated your suggestions and got the 3 paths (indicated as 3 white lines in script). The problem now is, I can't use the sweep2 component because there are 3 paths and 2 different rails. How do i get the smoth surface/breep that fits??

Attachments: