3d-Symbol distribution on a sphere object

[bjoerka]

I have a task of which i have no idea how to model it in a comfortable and better adjustable way.

I am trying to distribute a 3d symbol on a segment of a sphere. The target is to get a constant structure, similar like it is shown in the first image attached.

This was modeled in Maya or C4d by one of our collegues.

Now it is my task to bring this to a better stage to make a real planning to this.

 

My first attempt was to use the surface array command, but this leads into unregular gaps between the extruded tiles. (second image attached)

As i didn´t find any other tool i modeled it in steps.

Slicing the 8th of a sphere, extruded the outlines to nurbs curves, placed 3d loci on that curve and finally placed the 3d symbol according to the nurbs curve normal and rotated them to the center of the 8th of the sphere.

Final result - not what i expected .-)

 

Is this, expcept for a Marionette, impossible to get a better and more flexible workflow?

 

thanks!

bjoern

 

1st image: maya c4d design

 

2nd image - surface array

 

3d image - modeled manually

[Kevin Allen]

Surface Array Command

[bjoerka]

ii gave that a try, to be seen in the second screenshot, but at the lower end, the gaps are closed and the tiles are overlapping.

so, i want to find a solution where i can control the gaps from one to another element in a defined distance.

surface array doesn´t do that for me...

[Kevin Allen]

2 minutes ago, bjoerka said:

i can control the gaps from one to another element

not sure that's possible. Think about the math.

[bjoerka]

I think this is where Surface Array on a sphere is getting to it´s limit because auf the uv´s that are getting closer to each other as much as the top and lowest point of the sphere is reached.

sorry but i can´t explain it better.

it´s a bit like lattitude and longitude on our earth .-)

[Kevin Allen]

something would have to give, either the space between the objects or the sizes of the objects themselves.

[bjoerka]

you are right. but i think the problem is to calculate these positions on a surface which is double curved.

and the understanding of math, from my side, is as far in distance as Mos Eisley is....

[bcd]

Petri @islandmon we need you!

 

 

[VIRTUALENVIRONS]

I have a question.  This is probably related to Landscape stuff and symbols.  I don't have any knowledge of these modules.

 

CaB a 3D loci be converted to a symbol and that symbol converted to a primitive, circle or square.

Edited September 28, 2023 by VIRTUALENVIRONS

[Pat Stanford]

There are scripts that will allow you to replace a 3D Loci with an object or a symbol.

 

But I don't think they will help much in this case as none of the ones I have seen will do the automatic rotation that will be needed to get the "back" of the symbol facing the center of the sphere.

[VIRTUALENVIRONS]

Enclosed is a sphere built in VW's that is all 3D loci.   Perhaps someone can do something with it. 

3D SPHERE.vwx

 

 

[Jeff Prince]

10 hours ago, bjoerka said:

I have a task of which i have no idea how to model it in a comfortable and better adjustable way.

 

There are better tools for the job than Vectorworks.

You can do this in Revit, Rhino, or Blender with ease.

Of the 3, Rhino is probably the best bang for the buck for geometry, even considering Blender is free.

Rhino + Grasshopper can get you some nice parametric solutions to these problems, and there are already several solutions freely available for this task in particular.

[Benson Shaw]

@ all - I think two problems here:

1. The orig may be fudged a bit from the uniform tiling, if that's even expected.  See critique. But, no matter. . .

2. I think the tiles are laid out on great circles, rather than latitudes.  This allows even space and parallel edges along each course, rather than leaning in towards the pole, causing a tapered gap.  In a great circle arrangement, the courses have tapering separation (ie  meeting at the poles).  Tiles will overlap near the "poles", but overlaps can be deleted, leaving enough tiles to occupy required portion of a "sphere".

 

If the tiles are small/numerous, compared to the sphere radius, the widest part of the taper (between courses) can be somewhat small.  Easier for me to cover most of the ball, then rotate whole array to make the courses on angle. My tiles are pretty big, only 24 facets around the great circle.  40 or 50 facets would be closer to the orig example and allow more tiles/surface without overlaps.

 

Also, I approached this with poles on the layer plane. Second course is dupe of 1st, rotated around layer plane to align tile and gap center, then rotated about my (layer plane) polar axis.  Easier for me to think about it that way. Similar process could establish polar axis perpendicular to the layer plane.

 

-B

 

 

 

 

 

Edited September 29, 2023 by Benson Shaw
Opposing thumbs

[bjoerka]

Thank you Benson for explaining the way you did this.

I will try to follow this on wendsday... we have a holiday on tuesday in germany and i will not be in the office on monday .-)

But i think your idea is something that i can work with!

[Benson Shaw]

Yet another thought.  I can imagine another approach, but didn't try it yet, maybe not possible in vwx (See @Jeff Prince above):

 

A basket ball, a handful of Scrabble™ tiles. Hot glue.

Mark off a 1/4 or 1/8 area of the ball.

Heat up the glue gun . . .

• Stick a course of tiles across the marked off area. Follow a great circle so that they have desired gap and adjacent edges are parallel.
• Stick another course on either side of the 1st course.  Center these tiles on the gaps of the 1st course, but only a little gap between courses
• Repeat for more courses as desired.  Leave blank spaces per the OP design.

This scheme could not cover entire surface because overlaps. But would possibly work in the limited area without much deviation. Accept the deviation and move ahead.

 

The 1st course is great circle, but subsequent courses are laid neither on great circle, nor parallels. Each is a unique track. Possibly not even circular?

 

In Vectorworks this might mean moving and rotating all the subsequent tiles by "hand".  I will give it a try if I have a moment.

 

-B

Edited September 29, 2023 by Benson Shaw
Sky's the Limit

[VIRTUALENVIRONS]

I have been busy for the last few days, but had a little time tonight..  The spacing on these clones is little off, but that is only because it is for demo purposes.  If I had to do this for real, I could be quite accurate.

 

 

[Benson Shaw]

@VIRTUALENVIRONS Fun! but not addressing the requested conditions. 
I would be very interested to see your array modified to match the OP: —-*square nodes

*wide gap along rows, adjacent edge’s all parallel

*no (or very narrow) gap between columns. 

*each successive row offset to center on adjacent gap. ???

I think it might be possible in a small patch 

or else fudge the location/rotation of individual instances as needed

-B

[Benson Shaw]

Actually, didn't say those initial descriptions very well.  Rows and columns don't really apply.  Should be more like this:

 

Tiles are square extrudes, all same dimensions.

Within each course, tiles are equally spaced. Gap width is prox 1/4 tile tile width. Neighboring tile edges are parallel.

Adjacent courses are minimally separated. Tiles are centered on gap in neighbor course.

 

As stated by @Kevin Allen - Something has to give. 

I think it can come close (fudged) in a small patch, but not over whole sphere or even large areas.

-B

 

Edited October 1, 2023 by Benson Shaw

[VIRTUALENVIRONS]

Hi Benson,

How are things?  I have been a little busy last few days.  I will look at this again next time I get a hour or so.

regards...Paul

[VIRTUALENVIRONS]

I am not really grasping what this fellow wants, my limitation probably, not his or yours.  Is this something like what we are talking about.  Like the video, these are symbols and can be anything.

 

Paul

[BillW]

Not exactly what is wanted but if you can get a 3D surface, split how you want, which can be converted to a group of 3D polygons, I have a set of tools that can create offset panels. As a side note there are some interesting plugins in Blender which may be useful ie  Mesh: Geodesic Domes and Mesh:Tissue

 

I converted one of my plugins "Faceted Hemisphere" to a group of 3D polys and created the results shown below. It can take a bit of time to process but the advantage of the plugins is that the panel and shape settings can be changed. Note the normals of the 3D polys initially face inwards (I believe) however the faces can be reveresed before processing.

 

 

[VIRTUALENVIRONS]

Enclosed is the 3D model and a short video below showing what you can do with this model.  It may or may not be of help, but someone may find it useful.  In many ways it behaves like C4D Mograph although I don't think Mograph could completely encircle a Sphere is squares without some foreshortening.

The file is 2018 so almost anyone can use it.

FORUM FILE.vwxFORUM FILE.vwx

 

 

 

[bjoerka]

I didn´t reply for the last time - too many other things i have been working on ,.)

Thanks everybody for their input and inspirations!

We have done some work in Rhino and in vwx with grasshopper and marionette, but finally the job was done the traditional way .-)

cutting the 8th of a sphere in vertical by defined dimensions. curves extracted to duplicate 3d loci on these and then placing 3d symbols on the locus.

after the first design step the sub-construction was modeled with 3d polys, thickend, holes cut, plans drawn.

now it´s part of the structural engineer .-)

 

attached some screenshots of the (hopefully) final design...

 

 

Edited October 31, 2023 by bjoerka

[VIRTUALENVIRONS]

There is another thread going called "Geodesic Dome" that is right in line with this work.  There is a post on this,  on the main page below your post.

 

Nice work.

Edited October 31, 2023 by VIRTUALENVIRONS