[Gmsh] Re: Gmsh

[Christophe Geuzaine]

> Max Larson wrote:
> 
> Hi Christophe,
> 
> I recently downloaded the MS Windows version of your software and I
> must say that I am impressed.  I may be able to use it sometime in the
> future to generate ship's hull models for dynamic mooring analyses -
> I'll let you know how I make out if and when I try it.
> 
> I read your comment on your web page that the characteristic length
> can be "associated to another mesh (the background mesh)."  I would
> like to try using Gmsh to generate a computational grid for an ocean
> hydrodynamic FE model.  Typically, traditional software for this
> application first generates a TIN of usually poorly-shaped elements
> representing a best-fit to the real seabed and spot water depths. The
> software then generates a model grid that is somewhat independent of
> the original nodes such that the elements are nicely-shaped and sized
> to yield larger elements in deep water and smaller elements in shallow
> water.
> 
> Could you suggest how I might tackle this problem using Gmsh?
> 

There are three main possibilities :

- You can specify characteristic lengths at the points of your
geometrical model. Since the actual size of the elements is linearly
interpolated between these characteristic lengths, you may have to add
some "dummy" geometrical entities in your model in order to get the
desired element sizes.

- You can use "attractors" (see tutorial t7.geo).

- You can generate a background mesh, i.e. a scalar post-processing view
in which the values are the target element sizes. The problem with the
background mesh approach is that you need 1) a first (usually rough)
mesh and 2) a way to compute the target sizes on this mesh (this is
usually done through an error estimation procedure).


> Thank you in advance.
> 
> Sincerely,
> 
> Max
> ---------------------------------------------------------
> Max Larson, M.Sc., P.Eng.
> Senior Coastal Engineer
> Triton Consultants Ltd.
> 618 Thurston Terrace
> Port Moody  BC
> CANADA   V3H 4E4
> phone/fax: (604) 469-3563
> email: mrlarson at triton.ca
> URL: www.triton.ca

-- 
Christophe Geuzaine

Tel: 32 (0) 4 366 37 10    http://geuz.org
Fax: 32 (0) 4 366 29 10    mailto:Christophe.Geuzaine at ulg.ac.be