Development and Application of Parallel Agglomerated Multigrid Methods for Complex Geometries
نویسندگان
چکیده
We report further progress in the development of agglomerated multigrid techniques for fully unstructured grids in three dimensions. Following the previous studies that identified key elements to grid-independent multigrid convergence for a model equation, and that demonstrated impressive speed-up in single-processor computations for a model diffusion equation, inviscid flows, and Reynolds-averaged Navier-Stokes (RANS) simulations for realistic geometries, we now present a parallelized agglomerated multigrid technique for 3D complex geometries. We demonstrate a robust parallel fully-coarsened agglomerated multigrid technique for the Euler, the Navier-Stokes, and the RANS equations for 3D complex geometries, incorporating the following key developments: consistent and stable coarse-grid discretizations, a hierarchical agglomeration scheme, and line-agglomeration/relaxation using prismatic-cell discretizations in the highly-stretched grid regions. A significant speed-up in computer time over state-of-art large-scale computations is demonstrated for RANS simulations over 3D realistic geometries.
منابع مشابه
Application of the Schwarz-Christoffel Transformation in Solving Two-Dimensional Turbulent Flows in Complex Geometries
In this paper, two-dimensional turbulent flows in different and complex geometries are simulated by using an accurate grid generation method. In order to analyze the fluid flow, numerical solution of the continuity and Navier-Stokes equations are solved using CFD techniques. Considering the complexity of the physical geometry, conformal mapping is used to generate an orthogonal grid by means of...
متن کاملSpace-Time-Parallel Poroelasticity Simulation
The accurate, reliable and efficient numerical approximation of multi-physics processes in heterogeneous porous media with varying media coefficients that include fluid flow and structure interactions is of fundamental importance in energy, environmental, petroleum and biomedical engineering applications fields for instance. Important applications include subsurface compaction drive, carbon seq...
متن کاملUG-Interface for CAD Geometries
Numerical simulation of real-world problems in computer aided engineering often requires the processing of complex CAD geometries. Powerful tools with fast algorithms must be supplied in order to cope with the huge amount of data. As the software package UG, combining parallelism and adaptivity with multigrid methods on unstructured grids, represents the ideal platform for solving large-scale p...
متن کاملA Domain-Decomposed Multilevel Method for Adaptively Refined Cartesian Grids with Embedded Boundaries
Preliminary verification and validation of an efficient Euler solver for adaptively refined Carte-sian meshes with embedded boundaries is presented. The parallel, multilevel method makes use of a new on-the-fly parallel domain decomposition strategy based upon the use of space-filling curves, and automatically generates a sequence of coarse meshes for processing by the multigrid smoother. The c...
متن کاملParallel Smoothers for Matrix-based Multigrid Methods on Unstructured Meshes Using Multicore CPUs and GPUs
Multigrid methods are efficient and fast solvers for problems typically modeled by partial differential equations of elliptic type. For problems with complex geometries and local singularities stencil-type discrete operators on equidistant Cartesian grids need to be replaced by more flexible concepts for unstructured meshes in order to properly resolve all problem-inherent specifics and for mai...
متن کامل