[CIG-SEISMO] CIG-SEISMO Digest, Vol 69, Issue 6

[Dimitri Komatitsch]

Hi Dylan and Yingzi,

Just use the "define_external_model.f90" routine (which works with both 
the internal mesher and Gmsh/CUBIT), read your binary model into an 
array once and for all and send it to that routine and use it there and 
you should be all set.

PS regarding random models: beware that grid methods (including SEM, 
Discontinuous Galerkin, finite differences and so on) are usually not 
very good at that (if your random perturbations are big i.e. if you have 
sharp variations inside an element; in such a case it is often better to 
use methods designed for that (see e.g. the work of Randall LeVeque et 
al. http://faculty.washington.edu/rjl/ )

Best regards,

Dimitri.


On 10/30/2013 04:28 PM, Yingzi Ying wrote:
> Hi Dylan,
>
> I am in similar situation to do 2d simulations with some velocity
> structures, but with fluid-solid coupling...
>
> One solution to is to supply with "DATA/model_velocity.dat_input" which
> defines value in each GLL points. I tried this method but no success.
> PS. I have reported my problem here but no response...
>
> Another way is to give value of each spectral element, that you can
> define "nbmodels" and "nbregions" in "DATA/Par_file" with your own.
>
> If you have any better solution, please also let me know.
>
> Good luck,
> Yingzi
> On 10/30/2013 03:14 PM, cig-seismo-request at geodynamics.org wrote:
>> Send CIG-SEISMO mailing list submissions to
>>     cig-seismo at geodynamics.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>>     http://geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo
>> or, via email, send a message with subject or body 'help' to
>>     cig-seismo-request at geodynamics.org
>>
>> You can reach the person managing the list at
>>     cig-seismo-owner at geodynamics.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of CIG-SEISMO digest..."
>>
>>
>> Today's Topics:
>>
>>     1. SPECFEM2D - random velocity models and meshing (Dylan Mikesell)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Wed, 30 Oct 2013 10:42:51 -0400
>> From: Dylan Mikesell <mikesell at MIT.EDU>
>> To: cig-seismo at geodynamics.org
>> Subject: [CIG-SEISMO] SPECFEM2D - random velocity models and meshing
>> Message-ID:
>>     <CADgqVRb+3JFKBvTF7c5VChkhPsv7HKiNOLk-q78Gd=ULucOM8w at mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Dear SPECFEM2D group,
>>
>> I have a question regarding meshing and the EXAMPLES/INDUSTRIAL_FORMAT
>> example.
>>
>> I would like to simulate p-wave propagation with SPECFEM2D in the
>> attached
>> velocity model.
>> (the left plot shows the background velocity field, the right plot shows
>> the field after draping a random perturbation field over the
>> background. I
>> am using the right plot for simulation).
>>
>> I was trying to use/understand the INDUSTRIAL_FORMAT example so that I
>> could just drop this velocity model into place and go. I have a SeisUnix
>> velocity model. As far as I could tell, that example created a mesh with
>> xmeshfem2D and then mapped some given SEP velocity model onto that mesh
>> using a nearest neighbor routine. I can modify things to get an SU
>> model to
>> work, but I am not sure that is the way I need to go.
>>
>> Since I do not know too much about the meshing, I thought I would ask the
>> pros. Will the internal xmeshfem2D work for this velocity model or
>> should I
>> use something else? I see Gmsh in the 2D user manual and Cubit in the 3D
>> Cartesian user manual.
>>
>> The boundary conditions are not a big deal at the moment. I see there
>> is a
>> little extra work to get PMLs using the external meshers.
>>
>> Thank you for any thoughts.
>>
>> Dylan
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:
>> <http://geodynamics.org/pipermail/cig-seismo/attachments/20131030/c7ff00fe/attachment.html>
>>
>> -------------- next part --------------
>> A non-text attachment was scrubbed...
>> Name: rand_model.pdf
>> Type: application/pdf
>> Size: 2583634 bytes
>> Desc: not available
>> URL:
>> <http://geodynamics.org/pipermail/cig-seismo/attachments/20131030/c7ff00fe/attachment.pdf>
>>
>>
>> ------------------------------
>>
>> _______________________________________________
>> CIG-SEISMO mailing list
>> CIG-SEISMO at geodynamics.org
>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo
>>
>> End of CIG-SEISMO Digest, Vol 69, Issue 6
>> *****************************************
>>
>
> _______________________________________________
> CIG-SEISMO mailing list
> CIG-SEISMO at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo

-- 
Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France    http://komatitsch.free.fr