Energy Citations Database

Bibliographic Citation

 
Document PDF 741 K   View Document or Access Individual Pages  -   search, view and/or download individual pages
DOI 10.2172/763231
Title Elliptic Solvers for Adaptive Mesh Refinement Grids
Creator/Author Quinlan, D.J. ; Dendy, J.E., Jr. ; Shapira, Y.
Publication Date1999 Jun 03
OSTI IdentifierOSTI ID: 763231
Report Number(s)LA-UR-99-2673
DOE Contract NumberW-7405-ENG-36
Other Number(s)TRN: AH200103%%532
Resource TypeTechnical Report
Resource RelationOther Information: PBD: 3 Jun 1999
Research OrgLos Alamos National Lab., NM (US)
Sponsoring OrgM and A (US)
Subject99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; ALGORITHMS; LOS ALAMOS; OPTIMIZATION; PERFORMANCE
Description/Abstract We are developing multigrid methods that will efficiently solve elliptic problems with anisotropic and discontinuous coefficients on adaptive grids. The final product will be a library that provides for the simplified solution of such problems. This library will directly benefit the efforts of other Laboratory groups. The focus of this work is research on serial and parallel elliptic algorithms and the inclusion of our black-box multigrid techniques into this new setting. The approach applies the Los Alamos object-oriented class libraries that greatly simplify the development of serial and parallel adaptive mesh refinement applications. In the final year of this LDRD, we focused on putting the software together; in particular we completed the final AMR++ library, we wrote tutorials and manuals, and we built example applications. We implemented the Fast Adaptive Composite Grid method as the principal elliptic solver. We presented results at the Overset Grid Conference and other more AMR specific conferences. We worked on optimization of serial and parallel performance and published several papers on the details of this work. Performance remains an important issue and is the subject of continuing research work.
Country of PublicationUnited States
LanguageEnglish
FormatSize: 10 pages
AvailabilityOSTI as DE00763231
System Entry Date2007 Apr 16

Top