Create groups from clusters and align to cluster point

[Unaided]

Hello everybody!

I'm building a little ice compound to place a romb shape over polygons of a modified dodecahedron.

I start with next steps:
1. Create dodecahedron.
2. Triangulate polys with subdivisión X method.
3. Make a cluster with each center point of pentagons
4. Scale from dodecahedron center to spherize it.
5. Select and delete key edges to conform a spherical volume made with rhomboids.
7. I have ready the geometry to place objects.

The Ice compound procedure is creating copies from mesh and place copies on each center polygons.

Also add the array of kine.global of dodecahedron for translation and rotation, to 'sticky' the pieces on geometry support.

Now I need set orientation of each piece in local.Y axis. Because the rotation of each piece is different, I thinking to make groups of pieces throught cluster components and align a point cluster created in master piece (A) to a cluster point key in a dodecahedron (B).

I don't know how to make this.

Somebody have a idea?

[Unaided]

I attach two new screenshots to view that I need to do.
First picture show the state desired of each peace. I was rotated the components by hand over its local Y axis.
Second picture show the current state with ICE.

No one have any idea to make it this with ICE?

[Bullit]

It seems the 4 in forst picture are correctly perpendicular to the normal of polygons, or i am seeing it wrong?

See also if this helps. I tried to replicate but i must have made some mistake.

http://xsisupport.com/2011/04/11/ice-mo ... -to-edges/

[Unaided]

Yes, the five instances show it in the first image are aligned to polygon normals, with ICE (get polygon normal, and set with Convert Direction to rotation>Point At). This quads are not planar polygon, but I think the normal are computed anyway from average of its points.
However the Local.Y rotation is set by hand in the first picture on my previous post.

I think that the align to edges method I see from you link, may be useful for me. Instead I use copy from nbPolygons, I can copy instances from edge also. I try!

I have removed some edges (step 5 from gif sequence) to convert the dodecahedron in a spherical shape of rhomboids.

[Bullit]

If you move your dodecahedron you do have to add its global position. I wasn't able to get a correct rotation if i rotate the dodecahedron.

[dwigfor]

If you move your dodecahedron you do have to add its global position. I wasn't able to get a correct rotation if i rotate the dodecahedron.

Try get particle position & get particle orientation -> srt to matrix * (Multiply by) dodec.kine.global -> matrix to srt -> set particle position & set particle orientation.

Multiplication order matters with matrix. Make sure the "parent" global transform is plugged in second.

[Bullit]

This is a polymesh does it recognizes the copies as particles?

[dwigfor]

Oh my mistake.. Sorry I haven't really followed the thread and didn't realize it was ICE Modelling...

In that case, I think it's even easier. Take the local transform of each aligned copy * dodec.kine.global -> transform topo -> set topo

[dwigfor]

Able to follow notation? Did that work?

[Bullit]

No it didn't worked but i didn't had much time to get in it.

I got polymsh.CopyTopology_Transforms multiplied by dedoc kine global> transform topo, but after that the settopo gives me a context array issue. I have to get into arrays but i didn't had time except a feeble try of build array from set, select in array that also gave me problems.

[dwigfor]

Do you have a scene file I can check out (2013 plz)?

After glancing at pic at top of thread, I wonder if problem is adding the rotations. Can you try multiplying them there instead?

[Bullit]

The issue appears to be that each copy rotates by their own center instead of rotating by dodecahedron center.

I'll post it tomorrow.

[Bullit]

Here is the project.

ObjectToPolygon&Edges.rar