[Gmsh] dynamic remeshing

Edgar Last edgar-chat at live.com
Sat Feb 16 21:54:45 CET 2008


Zhan, thank you for your reply.

Your simulation is much more advanced than my attempts. At the moment the bubbles are not yet modelled to deform elastically. Instead they are rigid spheres. The acceleration of the spheres is calculated from the summed buoyancy and surface drag. Then I calculate a new mesh after each discrete translation of the bubbles, not efficient or accurate at all.

Do you have more information on Johan Hoffmans Solver? Is it part of gmsh?

Edgar
________________________________
> Date: Sat, 16 Feb 2008 09:53:39 -0800
> From: feng.hobble at gmail.com
> To: gmsh at geuz.org
> Subject: Re: [Gmsh] dynamic remeshing
> 
> Hi Edgar,
> 
> I am dealing with parachute simulation. It seems we shared same challenges.
> 
> Do you also have to deal with fluid-structure interaction? I guess so since the bubble wall can be regarded as a structure like airbag which is air-structure-air inside out.
> If so what kind of implement method are you using to solve the FSI problem?
> 
> Now I am trying to use the FSI Solver that's developed by Johan Hoffman to solve the highly coupled nonlinear systems. I also asked him about the local mesh update.
> He confirmed the capability of his package but I have not tried yet.
> 
> Zhan
> 
> 
> 
> Message: 3
> Date: Sat, 16 Feb 2008 10:12:16 +0100
> From: Edgar Last <edgar-chat at live.com>
> Subject: [Gmsh] dynamic remeshing
> To: <gmsh at geuz.org>
> Message-ID: <BLU140-W3383971C5BEAF0B2C9D814E5270 at phx.gbl>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> 
> Hello!
> 
> I want to simulate two (elastic) bubbles rising in a cylinder filled with water. The  tetrahedral mesh is moving together with the fluid which is dragged upward in the vicinity of the bubbles. Of course the mesh quality decreases during the process, resulting in highly irregular or even degenerated elements.
> 
> I am looking for a very fast remeshing algorithm that establishes a good mesh quality whenever it has become worse than some threshold. Is such a remeshing available in gmsh? Or some other open source 3D mesh tool?
> 
> What is the basic idea behind remeshing algorithms? Do they invalidate those elements that are quenched too irregular, maybe together with some more elements in their vicinity, and then a new mesh is created in the invalidated region? Or can such a recycling of most parts of the mesh not be performed efficiently, and the complete mesh has to be regenerated instead?
> 
> Please give a few links where the basics of remeshing are discussed on a not too sophisticated level.
> Thanks
> Edgar

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/