Skeleton fattener + mesh cage morph

#21

Blimey, it works!

1 Like
#22

Is arctic mode also causing the iso-curves to be displayed like that or are you using some sort of t-splines?

(Daniel Piker) #23

Ah yes, I changed some of the settings for Arctic in the display mode options - turning mesh wires on, and switching color & material usage to Rendering material. They are Rhino 6 subd objects - no t-splines involved.

2 Likes
#24

Gotcha, so I assume you’re doing the technique where you have to type the entire subd command in to get it to work.

(Daniel Piker) #25

Sort of, I made a button with a macro because I got tired of typing it so often.

#26

Nice…is this easy to do?

(Daniel Piker) #27

Yes - just right click an empty bit of the toolbar you want to add the button to, Edit Button>New,
Give it a name, and an icon, and in Command put:
subdfrommesh
enter

2 Likes
(Souderdesign) #28

So excited! Thanks Piker!!!

#29

fun fun fun! well done Daniel!

#30

i just woke up and saw this… BOOM

1 Like
(Giansbatta) #31

Beautiful work Daniel you are my hero!!

#32

Good work !

#33

Nice work, Daniel - these are great. Is the approach similar to B-Mesh?

(David Andres Leon) #34

This is awesome @DanielPiker !

Is there any chance to use fattener.dll as an external reference for a script (like K2)? Say for instance, I´m dealing with 1000+ nodes and I want to avoid calculating the spheres in the canvas.

#35

I really like the new Fatten command. Having worked with T-Splines extensively, I have missed this functionality for a while now… Also in conjunction with SubD in RhinoWIP… so stoked! Thank YOU!

(Abdullatz97) #36

Does it only work with straight curves? Because I can’t seem to make it work on curved curves. Amazing plugin though!

(Daniel Piker) #37

Yes, for curves you’d need to approximate them with polylines

(Michael Pryor) #38

Use the Curve to Polyline component.

(Gustavo Fontana) #39

hi @DanielPiker,

I’m trying to use Fatten component in the way I was using TsPipe from T-splines, and it’s not working for me.

Let me show you some examples:

  1. This is what fatten does to this simple lattice structure:


    GH file with internalised data here:
    fattener_pipes_flow_issues_01_gf_190210.gh (34.1 KB)
    here’s a video showing the poly flow/thickness/artifacts issues with it:
    https://www.dropbox.com/s/n5xytq1hy7mzjw8/inconsistent_pipes_fattener_01_gf_190210.mp4?dl=0

  2. This is a more uniform/desired output of pipe thicknesses using Dendro (I culled duplicates in my example so Dendro doesn’t get pissed, I know Fatten will ignore duplicates.


    GH file for that: dendro_pipes_01_gf_190210.gh (38.6 KB)
    and video: https://discourse-cdn-aws1.com/mcneel/uploads/default/original/3X/b/4/b45ea16507f5401a789a068abbf00edab5e8eabd.mp4

  3. And here’s the best result, using TsPipe in V5, what I had asked originally, when Bob mentioned I should check this work:


    GH file with the low-poly that Tsplines built: ts_pipe_mesh_01_gf_190210.gh (227.5 KB)
    and video of that old tool in Rhino V5 when it’s being setup (this includes a few pauses because it’s very slow, not interactive).
    https://www.dropbox.com/s/axcopypc5jew14s/Tspipe_pipes_01_gf_190210.mp4?dl=0
    You will notice in the mesh internalized .gh file here that they also have some mesh twisting, but it seems to still work ok for some reason.

I know in the documentation you mention that this has trouble with <4 curves coming together:

…but here the max is 4 at any convergence.

Also I think that the logic that I need to choose either pipe diameter OR sphere does not make a lot of sense for design flexibility.

I think pipe diameter is about what’s the minimum /nominal diameter at each curve’s midpoint. And the spheres should only dictate how large is the webbing between each curve. Does that make sense?

We’d really love to use this tool for a lot of controlled design structures and lattices. We don’t really like any of the lattice tool/plugins for GH, they are too constrictive and limited. But for us to build our own we need a good fatten tool. What you have is so close, but not quite there. Do you think this can improve some more?

Here’s some screenshots of fatten (blue) vs ts-pipe (green), notice how besides the thickness inconsistencies, there are also some weird normals/tears in the fatten mesh:


Happy to give you more examples of what works/doesn’t. I know it’s a hard problem.

Thanks,

Gustavo

1 Like
Chromodoris mesh and voxels
(Matt Gaydon) #40

Hi @DanielPiker,
I am working with @gustojunk on these tests and have been able to understand some of the issues in more detail.

Firstly the twisted meshes that create the inverted normals on the struts. I have been able to use the line input and planes to create spheres with better alignment but they midpoint ones still don’t seem to match the line controls or the plane. Is it possible to get more control over this alignment?


Secondary, Gustavo all ready mentioned it above briefly but the way the node connections are defined if one line in aligned through the centre of the other 3 the strut connects the side the line has gone through and does not build from the centre.

Clean Junctions

Broken Junctions

Will it be possible to make it so when a strut connecting to nodes together like in a lattice between the top and bottom layer is added the junction is subdivided down like with TsPipe and thus connecting thought the junction instead of sticking to one side. I can model up what I am thinking it would look like as a mesh if that helps.

Cheers
Matt

GH File used: 140219_fattener_pipes_flow_issues_RH-Forum.gh (42.6 KB)