[Trilinos-Users] Information on using ML for ERF

[Duncan van der Heul - EWI]

Hello Jonathan,

Thank you for your reply.

I just put 'complex' in the search field on the ML homepage  and found:

Search Results
Found 3 documents matching your query. Showing best matches first.
1.	Page ChangeLog
	Matches: complex(1)
2.	Member ML_Aggregate_Struct::max_neigh_already_selected
	Matches: complexity(1)
3.	Member MLAPI::MultiLevelAdaptiveSA::GetComplexity
	Matches: complexity(1)


And opened the first log file to find:

------------------------------------------------------------------------------
    
mystring = ml2_0_0_5

    * Changes in aggregation post-processing to improve aggregate quality.
    * Added wrappers so that ML can use Epetra objects & so that 
      AztecOO can use ML as a preconditioner.
    * Added block Maxwell capabilities for real block systems that are
      equivalent to the complex systems arising from Maxwell in the 
      frequency domain.

Sincerely,

Duncan van der Heul

dr.ir. Duncan R. van der Heul

Delft University of Technology
Faculty of Electrical Engineering, 
Mathematics and Computer Science
Delft Institute for Applied Mathematics 
Department of Applied Mathematics
Section Scientific Computing

Mekelweg 4 
2628 CD Delft
The Netherlands
Room: HB 07.230 
tel: +31(0)15 2782632
fax: +31(0)15 2787209
email: D.R.vanderHeul at tudelft.nl




-----Original Message-----
From: trilinos-users-bounces at software.sandia.gov on behalf of Jonathan Hu
Sent: Wed 9/3/2008 2:12 AM
To: trilinos-users at software.sandia.gov
Subject: Re: [Trilinos-Users] Information on using ML for ERF
 
Hello Duncan,

I'm happy to hear that you've had some success with ML. We tried to 
apply ML to equivalent real formulations to frequency domain problems 
but didn't come to any satisfactory conclusions, and so didn't release 
any code to support this functionality. Out of curiosity, what was the 
"update" that you found on the Trilinos web site?

Regards,
Jonathan Hu

> Date: Tue, 2 Sep 2008 17:46:41 +0200
> From: "Duncan van der Heul - EWI" <D.R.vanderHeul at tudelft.nl>
> Subject: [Trilinos-Users] Information on using ML for ERF
> To: trilinos-users at software.sandia.gov
>
> We are currently exploring the possibilities to use ML as a
> preconditioner to solve electromagnetic scattering problems.
> For purely real problems it works very good.
>
> However, we really would like to use it for complex valued
> problems.
>
> I found using the 'search function' on the ML site that there
> was an update to make it possible to use ML for solving
> an Equivalent Real-valued Formulation.
>
> Could anyone give me some pointers as to where we can find more
> information on how the use ML in this way?
>
> dr.ir. Duncan R. van der Heul
>   
-- 
Jonathan J. Hu, mailto:jhu at sandia.gov
Postal address: Sandia National Laboratories
                Mailstop 9159
                PO Box 969, Livermore, CA 94551-0969
Tel / Fax (925) 294-2931 

_______________________________________________
Trilinos-Users mailing list
Trilinos-Users at software.sandia.gov
http://software.sandia.gov/mailman/listinfo/trilinos-users


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://software.sandia.gov/mailman/private/trilinos-users/attachments/20080903/38bb66ff/attachment.html