|February 16, 2010, 23:53||
Join Date: Jan 2010
Posts: 13Rep Power: 6
Can anyone give a little advice?
In the long run, I want to model fluid-sediment dynamics around rock outcrops in rivers. That's in the long run, at the moment I'm trying to generate meshes automtically. Once I've done that, I'll move on to chosing solvers, and so on ...
I've spent the last week installing and using BRL-CAD to create an stl file that I can use in snappyHexMesh, and now I am speding another week learning how to use snappyHexMesh.
Does anyone know if this is the correct route to go for my type of application? How adaptable is this technique? I know one gets different types of meshes (tet, hex, poly ..), but I'm not sure when/if I'll need what, and if this method will be able to do that? Also other words are flying around, like: gridless, legragian, eulerian, ALE, unstructured, structured grids ...
I'm a bit concerned that I may be wasting my time.
|February 17, 2010, 12:21||
Join Date: Feb 2010
Posts: 1Rep Power: 0
I'm not familiar with snappyHexMesh but am intimately familiar with BRL-CAD's capabilities. What you're suggesting sounds like a functional workflow but will depend heavily on the complexity of your geometry and the quality of analysis you desire.
Ideally, you'll want to preserve an implicit or parametric geometry form until you're ready to generate a mesh in order to preserve shape and topology best. So going through STL is less-than-ideal as you've already lost geometric information by doing that conversion. Whether that even matters, though, depends on your geometry. snappyHexMesh could be perfectly sufficient without knowing more about your analysis goals.
BRL-CAD does have a converter to CUBIT, a cheap-but-not-free code from the Sandia National Labs that specializes in generating different types of finite element meshes. If I recall correctly, they're a little over $300 USD bucks, but well worth it if you have advanced meshing needs (they're one of the best). That conversion path will preserve the geometry in parametric form for most BRL-CAD entities and result in a very high-quality mesh of a specifiable type. That's a little more complicated arrangement to set up, obviously, since you'd have to get a license and hook up BRL-CAD to work with it, but an option if your problem calls for it.
That said, I would suggest figuring out what you're needs are by trying the more simple route that you've already figured out. When/if you figure out those needs and can be more specific on what the problems are, it should be easier to figure out a direction to go in.
Last edited by brlcad; February 18, 2010 at 02:27. Reason: updated cost
|Thread||Thread Starter||Forum||Replies||Last Post|
|Fluid dispersion into a free stream - Advice needed please||LSF||FLUENT||0||April 12, 2009 08:45|
|Hardware extension advice needed||Luk||CFX||7||February 2, 2009 18:05|
|SUPG for 2-D Navier-Stokes advice needed..||nicole||Main CFD Forum||2||November 23, 2002 00:06|
|advice needed!||Filip||Main CFD Forum||2||April 26, 2002 15:27|
|advice needed in solving linear system||yf yap||Main CFD Forum||5||February 5, 2001 00:21|