Alumni Project

Developing Interoperable Meshing and Discretization Technologies in the Terascale Simulation Tools and Technology (TSTT) Center

PIs: J. Glimm1,2, D. Brown3, L. Freitag4, Co-PIs: E. D’Azevedo5, P. Fischer6, P. Knupp4, X.L. Li2, M. Shephard7, H. Trease8, Affiliated Researchers: R. Armstrong4 (CCA), B Smith6 (TOPS)

1Brookhaven National Laboratory, 2State University of New York at Stony Brook, 3Lawrence Livermore National Laboratory, 4Sandia National Laboratories, 5Oak Ridge National Laboratory, 6Argonne National Laboratory, 7Rensselear Polytechnic Institute, 8Pacific Northwest National Laboratory

Summary

A major goal of the TSTT Center is to deliver interoperable discretization software for PDE-based terascale scientific simulation. Through both algorithm and software development, we are delivering high-order time and space discretizations and boundary conditions to application scientists for use in mesh-based simulation codes. SciDAC-enabled collaborations with the astrophysics, fusion, high-energy accelerator modeling, and climate modeling communities provide an early insertion path for our technology and assure the relevance of our interoperable discretization tool development.

Vision. Application scientists in many different areas can reach new levels of understanding through the use of high-fidelity calculations based on solving partial differential equations that model multiple coupled physical processes and multiple interacting physical scales. The optimal route to superior results in many application areas, and frequently the only way to obtain useful answers, is to use approaches that combine many different types of meshes and solution strategies into one simulation. Unfortunately, most modern meshing and discretization technologies are not interoperable making it extremely difficult to pursue these strategies. The Terascale Simulation Tools and Technologies (TSTT) Center recognizes this critical gap and is addressing the technical and human barriers preventing the effective use of powerful composite and hybrid methods.

TSTT technology development efforts fall into three broad categories: 1) the creation of a common mesh interface, 2) demonstrations of one-on-one tool interoperability, and 3) the development of new technologies to enable hybrid solution processes.

Common Mesh Interface Definition
There is a wide array of mesh and discretization strategies available to solve application problems and many times it is not clear a priori which is the best strategy for a particular simulation. See Fig 1. The only way to determine the proper choice is to experiment with a number of options. This is both time consuming and difficult because most mesh and discretization tools have very different programming interfaces. To enable this kind of experimentation, and as a first step toward interoperability, the TSTT team is developing a common software interface for its many mesh management infrastructures. A key aspect of our approach is that we do not enforce any particular data structure or implementation with our interfaces, only that certain questions about the mesh can be answered through calls to the interface. The challenges inherent in this type of effort include balancing performance of the interface with the flexibility needed to support a wide variety of mesh types. Further challenges arise when considering the support of many different scientific programming languages. This aspect is addressed through our joint work with the Center for Component Technologies for Terascale Simulation Science (CCTTSS) to provide language independent interfaces by using their SIDL/Babel technology. Implementation efforts of this specification are underway at most TSTT sites and additional efforts that use the interface to support interoperability in front tracking and mesh quality improvement algorithms are ongoing.

figure

A common interface would allow these three mesh types to be used interchangeably within a climate simulation.

One-to-One Tool Interoperability
To showcase the potential payoff for a software environment that supports interoperable meshing and discretization software, we have engaged in efforts that target one-to-one interoperability between TSTT tools. The most notable of these is the ongoing work to merge the FronTier front-tracking code with the Overture adaptive mesh management framework. The front-tracking methods developed within FronTier exactly follow the sharp interfaces between two different materials and have been used in a number of different scientific problems such as gas dynamics and petroleum reservoir studies. The adaptive mesh refinement techniques within Overture are used to insert more grid points only in regions where increased resolution is required. By combining the strengths of these two methods, we will develop a fundamentally new AMR front tracking algorithm that will simultaneously provide more accurate and computationally efficient simulations.

New Technology Development
In addition to combining existing technologies to enable hybrid solution strategies, new technology must also be developed. Our two primary efforts to date include the MESQUITE mesh quality improvement toolkit and the creation of a Discretization Library.

MESQUITE: To improve the quality of the hybrid meshes generated as part of the TSTT project, we are developing a freely available, comprehensive software package called MESQUITE that accommodates a number of different mesh element types, quality metrics, and state-of-the-art algorithms. We have completed the initial design of MESQUITE and a preliminary implementation which supports optimization-based node point movement schemes to improve two- and three-dimensional mixed element meshes on complex geometries.

The Discretization Library: The complexities of discretizing new applications on unstructured and adaptively evolving grids have hampered widespread usage of many powerful discretization tools. We are simplifying the development of application codes by creating a Discretization Library of mathematical operators and boundary conditions that are common to many applications. The TSTT center has a great deal of expertise in high-order and adaptive discretization strategies, but most of the implementations are tightly coupled to existing frameworks and are not suitable for direct insertion into a stand-alone library. Thus we are working to separate and re-implement low level operators from existing TSTT frameworks.

Future Plans. The TSTT center will continue development of new technologies to better enable hybrid solution strategies. Our common interface definition efforts will focus on the creation of interfaces that support mesh modification, parallel computing, and the interactions among the mesh, geometric domain, discretization library, and application field data. We will use newly developed technologies such as AMR front tracking techniques and MESQUITE to impact SciDAC applications now and as a showcase to highlight the promise of interoperable meshing and discretization. As interoperable TSTT technologies come on line, we will work with application scientists to develop new codes that use hybrid solution strategies to solve previously intractable physics problems.

Further information: http://www.tstt-scidac.org
James Glimm, Brookhaven National Laboratory,
Phone: 631-333-8155, glimm@bnl.gov
David L. Brown, Lawrence Livermore National Laboratory,
Phone: 925-424-3557, dbl@llnl.gov
Lori Freitag Diachin, Sandia National Laboratories,
Phone: 505-284-9711, ladiach@sandia.gov

back to project page

 

Home  |  ASCR  |  Contact Us  |  DOE disclaimer

URL http://www.scidac.gov/ASCR/ASCR_TSTT/reports/TSTT2003Annual2.html
Updated: Monday, 21-Aug-2006 13:31:20 EDT