attractor points from finite element analysis

Hi all - I have a definition here which takes data from an Algor analysis and distributes points along the edge curves of the analyzed surface so that they densify in areas of higher stress. I then drew curves on the surface between the edge points to create a sort of fibrous mesh across the surface.

The final bit of information I would like this definition to have built in is a way to make the drawn curves react to higher and lower stresses as they make their way across the surface. I imagine them being pulled toward zones of high stress.

I am thinking I would evaluate each surface curve for a list of points along their path and somehow make each of them move in some way as an average of all the stress points around them.

Not sure if that is the best way to approach this.

Can anyone offer any suggestions on how to get there?

(load the .csv file into the file path component at the very beginning of the definition)

Attachments:

over food court 1.3dm, 106 KB
over food court 1.csv, 4.1 MB
stress point driven_fibers on surface_4-20-11.ghx, 2.7 MB

Replies to This Discussion

Permalink Reply by Dedackelzucht on April 20, 2011 at 2:07am

If you could define the centers of the high stress areas? My file is only a idea how it could work.

Best regards

DeDackel

Attachments:

overfoodcourt1_Dackel.3dm, 112 KB
stresspointdriven_fibersonsurface_Dackel.ghx, 122 KB

Permalink Reply by Bennett Bossert on April 21, 2011 at 10:32pm

Thanks a lot. This is a great start to what I'm trying to achieve. I have manipulated it some and linked it up to the definition i've been using to generate the surface curves. i'm having a little problem thought that maybe you can help with?

I did not want the endpoints of the original interp crvs to move so I removed them from the data set to be translated. However, I cannot figure out how to add the endpoints back into the data set before drawing the new interp crvs. I believe it's just a matter of list management but I cannot riddle it out.

Any thoughts?

Thanks

Bennett

Attachments: