Grasshopper

algorithmic modeling for Rhino

Hey everyone,

in the course of some academic research project, i stumbled upon a problem many times discussed in this forum.

Rhino/Grasshopper should be used as a front-end to generate generic geometry files for CFD analysis of free flows around and object. To do so a bigger control volume is needed from which the object, to be to analysed, is subtracted. The result shall be exported as a single closed solid.

In this case the problem is asymmetrical and therefore can not be simplified by dividing it in half at an symmetric plane.

The GH component to do so should be "Solid Difference". But as the tab says "Intersect" and Rhino Boolean FAQ explain in further detail (in short: Intersect, Split, Join and Delete), it is just impossible to do so right now.

At the moment I use a workaround where all three solids are exported as different STEP files and the geometry module of the mesh program i use (Salome) is able do this Boolean intersection i talked about. It works in our process quit good but any other mesh program i have access to (Gmsh, enGrit) has no such feature.

I know that Boolean algorithms are a rather difficult subject and my programming skills are not sufficient to understand the ones Salome uses and write my own.

So my question is whether we can expect this feature within Rhino/GH anytime soon, like in Rhino 6 or probably never.

Greetings

Sven

Views: 993

Attachments:

Replies to This Discussion

In Rhino per definition a solid volume is a set of connected faces, where the normals of the faces all point towards outside. All solid operations assume that is the case.

You can "hack" your way around by either meshing the objects and just joining the results. Or you can use Nonmanifoldmerge as discussed here.

On the other hand, in reality, there cannot be any free floating but static objects. So why not fix the floating parts with some thin pipes?

Thank you Hannes for this information. I was not aware of the Rhino definition for solids. I'm afraid this answers my question whether we'll see this feature or not.

I just tried the NonmanifoildMerge command as well and it works fine in Rhino but exporting those geometries seams to be still experimental. The STEP file lacks in geometrical data while IGES 186 works but my mesh program can not read any volume information. Since the problem is often sitting in front of the computer I may have fetched wrong IGES entities. Nevertheless worth a try.

Like your pipe solution I gave different workarounds a try as wall before raising this topic. But those resulting in accumulation of high cell numbers in locations of the mesh which are unimportant for the solution and therefore only cost computational time. Hence the free floating version is the optimal solution and the workaround through a different software is alright. I was just wondering if i could be implemented in Rhino.

"there cannot be any free floating but static objects"

There can be any number of free floating objects computed with a steady-state solver for example. I'm thinking about an airplane or a submarine. But those have typically symmetrical planes when the angle of attack is zero and therefore no connection problem when cut in halves.

RSS

About

Translate

Search

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service