Dynamic Remeshing w/ regional edge length... Strategies?

Remesh_Plankton_circles_troubleshooting.gh

Hi out there,

Many thanks to D.P. and others for the recent wave of remeshing tools. I'm glad to report that they are already finding their way into our architecture practice.

I have a question for anyone who's been using or trying to use "Dynamic remeshing script with Plankton and Kangaroo" that Daniel shared on Jan. 8. I am trying to develop a surface into a dynamically remeshed mesh whose DUAL graph will be used as the base geometry for a large structure. The faces of the mesh will be sized from about 4 to 35 inches across, regionally.

I have had trouble getting the script to solve for a mesh without getting a couple iterations in and the simulation failing. I've attached a definition that shows where I am. I've been trying to isolate parameters that make it run more successfully but haven't found a reliable strategy yet. I included two different starting meshes to pull to, just to see if that made a difference. Changing the target sizes mid-simulation works "sometimes".

But does anyone have any kind of "best practices" for this exercise?

this was a "successful" run just before failing...

Comment

You need to be a member of Grasshopper to add comments!

Comment by Joshua Jordan on June 10, 2014 at 12:41pm: Thank you Daniel,

I noticed that thread right after posting. It seems like there is great work being done on the technique right as I look for it.

I've had a lot of success with MeshMachine, and still use your January c# script to introduce attraction locally. Most of my problems seem to arise when the falloff values are too high, e.g. too quick a transition between edge lengths.

But thank you as always and looking very forward to the next developments.

Comment by Daniel Piker on June 10, 2014 at 11:30am: Hi Joshua,

Since sharing that script in January I've found a number of ways to make the remeshing much more stable, particularly where boundaries are involved. These changes have now been incorporated into the 'MeshMachine' Kangaroo component.

I just updated this component the other day, with a way to vary the mesh density according to proximity to the boundary:

http://www.grasshopper3d.com/profiles/blogs/meshmachine-update?xg_s...

and as jjsolly asks for in the comments there - I will also add soon a way to control the mesh density with attractors like in that earlier script (or to use the colour of the input mesh to control the density).