An adaptive, three-dimensional, finite volume, Euler solver for distributed architectures using arbitary polyhedral cells
An adaptive, three-dimensional, finite volume, Euler solver for distributed architectures using arbitary polyhedral cells
The use of more than one cell topology in unstructured meshes may impose additional limitations upon the intrinsic adaptivity of the mesh. Additional geometrical entities have been used to install descriptions of arbitrary polyhedra within an unstructured mesh, permitting the incorporation of any mix of cell topologies, whilst maintaining the intrinsic of the mesh.
Three-dimensional, unstructured, hybrid meshes are constructed using a modified multi-block approach in which unstructured data is used to naturally incorporate mesh singularities, and permit zonal adaptations. The meshes are used to demonstrate a finite volume Euler flow solver, which efficiently operates on an edge structure such that no limitations are imposed upon the mesh topology.
Results obtained using a first order upwind scheme demonstrate the shock capturing abilities for a range of two and three-dimensional transonic flows. The implementation of a higher order method, using the MUSCL formulae, illustrates the non-trivial nature of applying such techniques within an arbitrary cell environment, whilst solutions obtained for two and three-dimensional transonic flows demonstrate the increased resolution obtained.
The implementation across distributed platforms is described in detail, with good performance results presented for a range of architectures, including a workstation cluster and an IBM SP2.
An adaptive mesh algorithm is employed to automatically identify and resolve local flow features. No limitations are placed on the adaptation of mesh cells, which is demonstrated for a supersonic internal channel flow, where the strong shock waves are clearly captured using a hexahedral to polyhedral strategy.
University of Southampton
Rycroft, Noel Christopher
3b44b366-42da-4fd6-8112-25f3005b0dd4
1998
Rycroft, Noel Christopher
3b44b366-42da-4fd6-8112-25f3005b0dd4
Rycroft, Noel Christopher
(1998)
An adaptive, three-dimensional, finite volume, Euler solver for distributed architectures using arbitary polyhedral cells.
University of Southampton, Doctoral Thesis.
Record type:
Thesis
(Doctoral)
Abstract
The use of more than one cell topology in unstructured meshes may impose additional limitations upon the intrinsic adaptivity of the mesh. Additional geometrical entities have been used to install descriptions of arbitrary polyhedra within an unstructured mesh, permitting the incorporation of any mix of cell topologies, whilst maintaining the intrinsic of the mesh.
Three-dimensional, unstructured, hybrid meshes are constructed using a modified multi-block approach in which unstructured data is used to naturally incorporate mesh singularities, and permit zonal adaptations. The meshes are used to demonstrate a finite volume Euler flow solver, which efficiently operates on an edge structure such that no limitations are imposed upon the mesh topology.
Results obtained using a first order upwind scheme demonstrate the shock capturing abilities for a range of two and three-dimensional transonic flows. The implementation of a higher order method, using the MUSCL formulae, illustrates the non-trivial nature of applying such techniques within an arbitrary cell environment, whilst solutions obtained for two and three-dimensional transonic flows demonstrate the increased resolution obtained.
The implementation across distributed platforms is described in detail, with good performance results presented for a range of architectures, including a workstation cluster and an IBM SP2.
An adaptive mesh algorithm is employed to automatically identify and resolve local flow features. No limitations are placed on the adaptation of mesh cells, which is demonstrated for a supersonic internal channel flow, where the strong shock waves are clearly captured using a hexahedral to polyhedral strategy.
This record has no associated files available for download.
More information
Published date: 1998
Identifiers
Local EPrints ID: 463498
URI: http://eprints.soton.ac.uk/id/eprint/463498
PURE UUID: 0968e372-d5f2-4de0-a192-848fe6cb7cbd
Catalogue record
Date deposited: 04 Jul 2022 20:52
Last modified: 18 Apr 2023 16:30
Export record
Contributors
Author:
Noel Christopher Rycroft
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics